I've come to the conclusion that weigh-sorting empty brass (of a given headstamp) is a worthless exercise. I believe that case volume consistency is more important and am working on a test.
Just recently, I weighed up 50 rounds of once-fired R-P .270 WIN cases, after being FL sized, trimmed to length, primer pockets uniformed, and necks outside turned for consistency. The lightest was 197.4gn, the heaviest was 202.9gn, for a 5.5gn spread. I then weighed all the cases after being overfilled with RL-22 and struck level and got the max powder capacity. The max difference was 1.7gn. Avg was 61.9 w/.373 Std. Dev. The most frequent charge wt. (mode) was 61.7gn. I ran a computer sort and picked 2 boxes of 20 rounds with average case capacities of 61.7 and 62.1gn. with 0.13 Std. Dev. per box.
20 are loaded to the exact same cartridge base-ogive length with the exact same powder weight. I intend to chrono groups of bullets that have been either: weigh sorted light/heavy, base-ogive sorted shorter/longer, or completely random wt & base-ogive. These will be 5-round groups each. Probably not statistically adequate, but that's what I'm giving it. I'll hopefully report back in a few weeks, since I've already had life interruptions to my range time 2 weekends running.