Assuming that none of the loads you're using are anywhere close to going subsonic at 300 yd, there are a few things you can probably rule out. You're describing going from 3/4 MOA at 200 yd to 2 MOA at 300 yd. That is a huge difference and is likely not an issue with the rifle itself or the load. A rifle and load that is shooting sub-MOA at 200 yd doesn't suddenly fall apart and turn into a precision shotgun in only 100 yds farther. In the absence of some external factors, that would almost defy the laws of physics, with one possible exception. What you're describing sounds most like either a shooter-based issue, or possibly one of a few possible external factors.
If shooter-based, my guess would most likely be something you're ding different at the longer distance. Hard to say "what" exactly, but something significantly different to cause that big of a difference in precision. It might be positional, or it might be related to the optic. Best way to rule out shooter-induced error is to have someone else you know can shoot to try it out. If they get the exact same results as at 200/300 yd as you, it would point to some external factor.
If external, I'd put at least a flag or two in the 200-300 yd region. Even 75 gr .224 bullets can be pushed around a lot by wind past about 200 yd or so, particularly at the relatively slow velocity you're likely getting out of an AR with a fairly short barrel. If not for the fact that you also get this effect with 75s, I'd attribute it solely to some wind and very light .224 bullets (<60 gr). I shoot 80.5 gr and 90 gr .224 bullets regularly in F-TR matches. Even shot out of 30" barreled bolt guns at very respectable velocities (2850+ fps), the 80.5s moved noticeably farther offline at 300 yd than the 90s, even in relatively mild wind conditions. I've shot plenty of Hornady 75s at 300 yd from an 18"-barreled Sig 556 DMR rifle and you really have to be on top of the wind conditions to keep the groups small. The lighter the bullet, the worse the effect will usually be.
It's also possible that your optic is part of the issue, even though it sounds like you've payed close attention to that. What power does your scope go up to? Are you running it at maximum mag at all distances? Again, letting someone else whose shooting skills you trust try it out may be the easiest way to spot something out of spec. The possible exception I mentioned above would be if you're getting wild velocity excursions for some reason. If the velocity excursions were pretty large, which although may not be typical, is certainly not outside the realm of possibility with a semi-auto, it could possibly be a contributing factor to the poor precision at 300 yd, but have less of an impact at 100/200 yd. Frankly, the velocity variance would have to be extremely large to account for this, but it is possible, and is something you can directly measure, even if only to rule it out as a cause.
I think what I would probably do first in your shoes would be to very carefully shoot multiple groups (each) at 100, 200, and 300 yd, using only the 75 gr load (i.e. heaviest bullet possible). I'd do it all at the same time, under the most consistent and benign conditions I could, so as to remove as much potential wind effect as possible. I would try to maintain the best shooting fundamentals possible, and closely monitor the conditions while shooting. Then I would very carefully measure the groups, determine average extreme spread, and compare the averages of several groups at each distance to make absolutely certain the precision difference between 200 and 300 yds is really as bad as you're describing. It's very easy to gain the perception that what you described is happening, especially when shooting a variety of different ammunition at different distances, when in reality it is the shooter, using light bullets in windy conditions, or some easily explained phenomenon. However, rigorously demonstrating it is actually happening requires consistent testing conditions, using the same ammo, etc. If the breakdown in precision between 200 and 300 yd still holds true following such a test, then I'd try to get someone else behind the rifle to see if they obtain the same results. If necessary, at that point it may be necessary to isolate one single variable at a time to try and determine what is going on.