I gotta agree with Hammer here. It seems like the Ki-84 does have reliability issues built into its model.
The KI-84 shedding parts at high speed I always figured was a bona fide design flaw -- that is, insufficient structural resilience in the design itself, rather than modeling the poor quality materials. I could be wrong on that point, though.
But the 37mm Il-2 recoil issue has nothing to do with quality control or materials. Great breakdown there, but I'm not sure it's applicable to this particular case (though I totally agree that jamming has no place in AH). Instead, the recoil effect occurs just because it's a big gun firing a big shell, and because the two big guns were not synchronized. There was not a failure of a synchronization mechanism that we're leaving out in AH because there never was one.
So how do you account for that?
If I understand the situation correctly, there was no synchonization "mechanism" that kept the guns firing at the same time. But they are two guns that should, in theory, be firing at the same rate if they were constructed exactly the same. Pull the trigger, both fire. Keep the trigger depressed, and they would continue to fire at the same time if the guns were perfectly identical. So, in RL they went out of sync why? Because in RL there are variabilities in the guns' mechanically limited firing rates, even though they were not designed to fire any differently.
So, I switch the question back to you. If you are modeling a game and want to be consistent with all weapons, what kind of mechanical variability in the firing time do you model in for each gun? .1 second? .3 second? .01 second? How do you even source that kind of information?
If you are going to model a variable rate of fire for a plane carrying two guns, shouldn't it be modeled for planes carrying only one gun? Shouldn't there be a variability in the rate of fire that would affect synchronized guns if one of them just happened to be inadvertentely built to fire slower?
Again, at the end of the day, IMO you would be unnecessarily complicating matters to take it to this degree because one plane seems to be benefitting from standardization moreso than other models.