5.8 gr variance on a case that weighs 190 gr is right at 3% -- But brass is much much heavier than powder, around 10 fold, so the powder capacity variation is more like 0.3% Lets say a ball park 7-08 load has 40 grains of powder. 40 x .003 = 0.12 gr. You think that much variation in powder will matter in an AR-10?
I've had lots of Lapua brass have several (3-5) gr variation too. I read somewhere that most of the variation in weight is in the base -- the extraction groove. So actual case capacity is less related to weight that I used to think.
In summary -- proof is in the shooting. Anneal that used brass, clean it up good and FL resize and tell us how it shoots. I am mostly a bolt gun shooter, so I could not tell you how to really reload for a AR-10 and tell if the brass is the problem in an accuracy matter. Seems to me the inherent reduction in accuracy with the autoloader makes quibbling about brass less of an issue -- but maybe some of you can get to 0.5 MOA with your guns???
Do I KNOW that some brass is better than others. Oh heck yes. But weight variation seems to dimm in terms of the way to check that. I think crappy brass has poor quality control of the brass metal itself and poor quality control of neck thickness. If you have variation of either, no amount of annealing with even the very best annealer will fix the issue. Neck turning is just a pain -- and with an autoloader, you shoot more by definition. Not an option. And if there is variation in the brass itself, I mean the composition of the metal, you obviously cannot fix that. That is why we discuss brass being "too hard" or "too soft" -- but what if a crappy manufacturer has hard and soft all mixed in?