A bit of background where I'm coming from here:
I was originally drawn to games like Red Orchestra and Insurgency back in the day of them being mods specifically because they attempted to do the best job I had ever seen in a game up to that point of trying to simulate what it's like to to actually handle and fight with these weapons.
However, as I've learned more over the years about what actually is entailed in shooting firearms, I'm somewhat shocked that the developers have not sought to amp up the fidelity of the gun simulation to account for some pretty glaring errors in the way it currently works. I mean, some of this stuff you could just find out by browsing youtube videos about the guns, much less talking to people who actually use them.
Some of the biggest errors and oversights:
No heat management. There's a reason the SAW was developed with a quick change barrel. Red Orchestra incorporated overheat and barrel changes into their MGs over a decade ago. Heat management also extends to other areas of the game. All rifles will experience varying degrees of degrading or shifting accuracy as the barrel heats up. If the heat buildup gets worse you'll have more malfunctions or changes in firing speed. If it gets bad enough with mag dump after mag dump then it will be destroyed and unusable. That's why you don't use assault rifles as machine guns.
Furthermore, there's a reason you never see suppressors on MGs. The heat buildup in a silencer is massive. Sustained firing will cause it to fail and blow up. It will also probably result in more heat buildup in the barrel or in the case of ARs (because of their gas system design) more buildup of heat in the receiver.
This all needs to be modeled to properly differentiate the ups and downs of the various platforms and accessories. As of right now there's actually too little reason not to always use a suppressor if you can. And there's no reason not to dump mag after mag repeatedly without concern for the consequences to the weapon.
---Heavier barrels will also be less effected by heat. But the weight penalty of a heavier barrel doesn't seem to be properly modeled.
---Certain gun designs (like being open bolt), and handguard or heatshield designs, will also do a better job of dissipating heat. So this is another way that the differences between guns can and should be modeled.
Overall equipment weight doesn't have enough of an impact on performance.
There's a reason that US soldiers actually say they wouldn't want to wear more armor than they currently do, despite the added protection, because they are concerned that the lack of mobility that comes from that would be more likely to get them killed by not letting them be nimble or fast enough in combat. There's a reason soldiers don't want to pack around the weight of a full combat loadout of 7.62x51 ammo, preferring the 5.56 weapons for that reason. There's a reason soldiers historically hate being the ones who have the carry around these heavy squad support weapons despite the weapon being more effective in many roles over the standard issue rifle.
You need an endurance meter that is effected by overall weight and endurance is depleted with actions and movement and to a small extent firing (based on the gun used and the position it's used from). Endurance levels effect aimsway and recoil management to a degree. It can also effect sprint speed or sprint duration.
Overall weight should also effect your transition speed between positions and transition time into sprinting.
Weapon weight and balance seems to not be taken into account with weapons handling and performance. Heavier weapons have drawbacks in reality that aren't properly represented in the game. If this were modeled properly then you wouldn't see the need for artificially making the SAW so inaccurate as the only way to make everyone not want to use a SAW over an M4. A heavier weapon should take longer to swing from one side to another. It should take longer to bring the sights on target. You should lose more endurance the longer you're moving with it or holding it up to aim down it's sights. It will likely also impede your ability to transition between positions and sprint if you're holding the weapon in your hands as opposed to slung around you.
Adding accessories to the gun will increase overall weight which should increase these problems, but it also will aid in reducing recoil due to the increased mass of the weapon.
As for balance, you need to represent the inherent disadvantages that exist for tricking out your gun with accessories at the front end of the muzzle or handguard. The more muzzle heavy your gun gets the harder it is to swing around quickly and accurately to adjust your aim. Also, the more endurance you will lose as you try to hold out that heavy long gun, which can translate to more aimsway. Although this can also have beneficial effects on recoil reduction. Being too back heavy as in some bullpup designs can make the gun more difficult to control in recoil.
Having a gun that is too long can also hinder your agility around corners and windows.
Body armor. Now I can't be 100% sure, but my impression is that body armor is not properly simulating how body armor works. It seems to be treated as just increasing hit points. But this is not accurate. Certain kinds of armor plates can block pistol rounds all day long without any degradation to the plate or even injury to the person being hit. Heavier types of body armor can block even 5.56 rounds unless it's armor piercing. Some types of body armor will experience degradation to the integrity of the plate from successive shots, and this should also be modeled. Some shots are so powerful that they can cause blunt injury to the person being hit even if the round is stopped, but not much damage in the scheme of things compared with a bullet wound. However, even non-lethal hits against armor can have the effect of stunning the person hit or causing flinching from pain, and therefore delaying or hindering their ability to respond effective effective return fire, depending on the type of round or the type of armor. Body armor should be physically represented as plates on specific areas of the body which, when struck, result in blocking shots outright or mitigating some of the energy.
Being forced to make compromises in weapon assembly that make no sense logically. There is nothing stopping me from having both a heavy barrel and a suppressor on my weapon. There is also nothing stopping me from having both a laser light and a flashlight. It's not necessary from a gameplay standpoint either when you have a points system for gear. Doing a better job of modeling the downsides of these using these various accessories, or too many accessories at once, would obviate the need you feel to arbitrarily restrict what players can put on their weapons.
Night vision illuminators and pointers. You don't simulator the downside of nightvision which is the difficulty of using your sights through them. That is why they have IR pointers that allow them to see where their gun is aimed so they don't need to look down the sights. Some optics do better than others at being viewed through nightvision. It probably makes for longer sighting time and less accuracy on ironsights. The illuminator is used like a flashlight, but has the benefit of nobody being able to see the light without nightvision because it's in the infrared spectrum.
Not enough fidelity in ammo modeling. You can do better than the basic "better against fleshy targets" or "better against armored targets". Get some data on specific ammo types and how they differ.
For instance, armor piercing rounds will have more recoil kick because you're throwing a heavier bullet at high velocity. They might also have more bullet drop at longer ranges.
In contrast, a 5.56 green tip round might not behave much differently from an FMJ round in terms of recoil and ballistics, but you only gain a slight increase in penetration effectiveness (It's not true armor piercing), but you actually get much worse stopping power against flesh as the trade off. That's why green tip is regarded as a bad round. It was designed to let you shoot more effectively through car windows but it does just about everything else worse.
These are only a few examples of how you're missing out on the potential of modeling completely different tactical options with both upsides and downsides by not delving into more specific ballistics profiles. More variety in ammo types would add both customization fun and realistic fidelity.
Understanding this freely available ballistics data is important to accurately simulating the performance of an M16 vs an M4. For instance; the 5.56 loses terminal ballistics effectiveness (stopping power) rapidly as you get away from an 18-20 inch barrel. By the time you get down to a 10 inch barrel you start to run into horrendously poor performance. This is because the 5.56 depends on velocity to be deadly but you lose so much velocity with a 10 inch barrel that your effective killing range for the 5.56 becomes very short. This is why the military is looking at other caliber options currently (like 6.8 caliber). Although changing the ammo you use can mitigate a lot of these problems in the 5.56, such as using the 77 grain SMK 5.56, you still run into significant problems when you go down to a 10 inch barrel (the severity of the problem is just decreased).
In contast, the 7.62x39 won't suffer so much in terms of killing power when you put it into a shorter barrel because it depends less on velocity to be effective (due to being a larger bullet, having more mass). It's important to understand the nuances of how one round behaves differently than another because if you just assume they all react the same to different barrel lengths then you won't get an accurate simulation of how the weapons differ in handling and performance.
The point being: there are significant enough differences between round types, and between different weapon platforms the rounds are used in, that it deserves to be modeled in specifics based on ballistics data. It would make a significant difference on how the various weapons play out in combat.
Suppressors do effect weapon performance in reality so they should in-game. Recoil will be reduced by slowing down the expansion and release of the gas out of the barrel over a longer period of time, thus resulting in less sharp spikes of recoil. As well as the effects that having more mass on the end of the gun will always reduce recoil further. They also give extremely minor gains in velocity and sometimes can increase accuracy.
Flash hiders. I don't know if you model their positive effects on weapons so it is noticeable when you don't have them equipped. A flash hider will save you from losing your night vision when you fire and probably not interfere with night vision scopes as much. A compensator can make seeing your target in fully automatic fire difficult or make follow up shots more difficult from a user's standpoint due to the concussive blast distracting you and flash of light obscuring your vision - although it depends on the weapon and the ammo used. Short barrels with high velocity rounds will suffer the worst effects by having more unspent powder coming out of the end of the gun in the form of a larger explosive flash.
Large caliber weapons cannot be effectively suppressed even with a suppressor. There is too much energy for it to be made movie-quiet. Suppressors on smaller caliber weapons also shouldn't be movie-quiet when you're still using supersonic ammo. The bullet makes a crack as it breaks the sound barrier. A suppressor will still help obscure where the bullet came from, but we should have the option of using subsonic ammo to achieve better quietness - but this comes at the cost of a lot of velocity (ie, power, range, and penetration lost).
Some weapons like the 5.56 also do very poorly when suppressed, forcing you to turn the gas system up higher to ensure the gun operates reliably. And having more gas go into operating the bolt will increase your recoil, if that recoil force is not offset by the advantages of the suppressor.
You can't actually shoot subsonic ammo in 5.56 and have the AR platform cycle reliably, which is why 300 blackout was created as a cartridge designed to be fired subsonic from the start.
Malfunctions. America's Army got this right in their game about the need to incorporate a button used to clear the gun of malfunctions. It's a very significant part of realistic weapons handling and a reality that will impact combat. It's part of whats separates a genuine shooting sim from an arcade shooter.
It also is important to model this if you want to bring about the realistic differences between certain weapon platforms in terms of their inherent reliability, differences in ammo reliability (where it exists), how well they handle the fouling caused by continued firing, and how well they handle heat buildup, and how easy the gun is to clear from malfunctions in the first place (some weapons are faster to clear and some weapons malfunction in worse ways than others).
Some weapons will malfunction less to begin with. Some variants of the AR will malfunction less than others. Some weapons are easier and faster to clear of malfunctions than others.
Shooting suppressed will cause some guns to get dirty faster and causes more wear on the gun's internal parts. Some guns perform more poorly suppressed than others to begin with, in terms of reliability.
Overheating your gun should lead to more frequent malfunctions and in extreme cases total failure of the gun.
Vertical grips don't always benefit the handling of a gun. A vertical grip on a standard length shotgun, without even having a pistol grip attached to it, would be largely useless and might even be ergonomically worse for handling. Typically the only reason you want to use a vertical foreward grip is if you're using a platform that has a very short barrel, because then your front hand ends up resting at a more natural angle. But, if you have a longer barrel, you want to move your hand forward more to benefit from better recoil control and aimsway control, which then makes the vertical grip angle feel unnatural and worse. There are other types of grips that competition shooters use like angled grips or side grips because they make more sense with longer barreled rifles.
You also have to factor in that vertical grips increase control in certain contexts more than others. For instance, with a short barreled M4 they will give you good control over the gun in full auto and be more comfortable to hold over the long term in a raised position (lowering endurance loss). However, for controlling the gun in repeated single shots, the vertical grip may provide no significant advantage.
More optics options. There should be the option to have a red dot style optic with a flip mounted magnifer that can be flipped into place when you want more range, and then taken down when you want close quarters combat. There are also different types of reticles and scope qualities that could be represented which all have different advantages. Different types of iron sights have different advantages too.
There are a lot more options that could be incorporated into customizing some guns, especially the AR platforms. Although I don't know how practical a lot of this stuff is for a general infantry squad, you might still see special forces taking more liberty with their gun customization.
Different triggers would be an example of this, something which competition shooters do a lot of. It can impact semi auto accuracy in providing a more smooth and quick trigger pull that is less likely to knock the gun off target, and follow up shots will have better speed and accuracy. It probably doesn't have significant impact on full auto.
But that does also bring up another point: Which is that you have to factor in trigger quality when determining how different guns will handle. This can be a big deal in a weapon's single shot accuracy to consider when modeling the gun's performance.