Two points. First, my mind wandered a bit on the post and I got to just addressing wanting to know the absolute case capacity. If you aren't using QuickLOAD, that's less directly helpful, but not zero help. Look at
this thread from a couple of days ago about case capacity to get the numbers on the significance of weight difference. Bottom line, with 70:30 cartridge brass having a specific gravity of 8.53, it works out for most rifle loads that it takes about 1.4 grains difference in brass weight to have the same effect on pressure as changing the powder charge weight 0.1 grains.
From the above, if you have a load that is insensitive to charge weight changes over a span of 1.0 grains, and you throw charges over a span of half a grain, then it will take a brass weight difference that exceeds 7 grains to drag the most extreme charges you throw into opening your groups up. Loads that insensitive are not uncommon, but I've had a couple over the years that were much more forgiving (2.5 grain span for one .308 load of Brigadier 3032 uisng the 155 grain Sierra Palma bullet) and less (.4 grains for another load of AA2520 with the 168 grian SMK). This echoes some of Hatcher's results with coarse and fine grained powder from developing National Match loads, the latter being touchier. I considered the latter unhelpful as it didn't have good immunity to temperature change and was picky about brass capacity uniformity, too, of course. The former was immune to all conditions I could throw at it. It was great, so they up and burned the Brigadier powder plant in Scotland down to be sure I couldn't get any more.
Dan Newberry's
OCW site is devoted to identifying forgiving loads, if you want to read about doing that?
If you are just sorting brass within a given lot, and it all has the same load history, weight comparison likely tells you all you need to know about capacity for all but bench rest match purposes, and you probably only have to sort it once in the life of the brass. Just be aware cases grow capacity over time due to trimming, which weight change catches, and head brass migrates rearward, which weight change does not catch, and which is not reliably uniform. Denton Bramwell's
article on why case head expansion and pressure ring expansion are not a reliable pressure measures shows this indirectly. He measured head expansion happening sometimes at very low pressures, like around 32 kpsi for some cases, and not until 65 kpsi for other cases all from the same lot and with the same load history. Likewise, he shows pressure ring expansion erratic. What this means, and slicing enough cases will show this to you, is some cases have more of their brass in their heads than others or else have it more on one side than others, allowing, as Rocky said, more expansion under pressure due to the resulting thin places.
I don't recall reading of anyone doing repeat loading of cases to see whether muzzle velocity variation for a load tended to follow particular cases or not? That is, whether you could identify cases within a lot that tended to produce the same MV based on how they expand? It would clearly be of no consequence at shorter ranges to most kinds of shooting—say to 300 yards and under—but the 1000 yard and bench rest shooters would surely like to know?
I recall Merrill Martin reported finding one case in an otherwise matched set that always shot slightly lower than the others (this was bench rest shooting). He measured carefully and found it to be the only dimensionally perfectly uniform case in the batch. The others were all thinner on one side than the other and presumably expanded a little more due to the thin portion, thus lowering barrel time slightly. That apparently let the muzzle move up a little higher by the point of departure. I don't recall him reporting velocities.
So, fussy is as fussy does, but some of these tiny changes actually do have measurable effect, at least in some loads in some guns. If you are in a situation where it is enough to matter, then you want to know about it. As always, the final bottom line is testing in your own gun under your own shooting conditions to see whether any effect is apparent or not?
I am not a fan of blowing off precision with the declaration that "the gun out shoots me as it is". I've seen too many beginners in service rifle match shooting say the same, only to find their scores jump up ten points when someone lends them a match rifle. You only have to turn ten almost-makes-its into scratch hits of the next higher scoring ring for that to occur. So, I'd rather err on the side of precision. I can prove statistically that it takes far less time and fewer rounds to first overdo exactness, then eliminate each required step until you isolate score affecting factors, than by gradual incorporation of such steps. But, as always, the key is your purposes. A deer will most often be less sensitive to that short difference between an almost-makes and a scratch hit than a scorekeeper is.