These are exactly my thoughts. Many say that the case should be trapped in a sizing die to avoid any variance from spring back.
I wonder if the rear of the case, the part that "shouldn't" expand, may also contribute to differing velocities. As, maybe, individual cases, which weigh the same, expand to differing rates at the rear of the chamber. Changing the specific volume of the chamber.
I wonder if maybe one might drop a case into a measured cylinder and see how much water is displaced by the case. This could tell how much volume is taken up in the chamber.
And, as was said earlier, yes. Water weighing for internal volume
is the most precise. It is also quite laborious. After doing some weighing with and without water, sorting bullets and grouping heavy cases, light cases and exact weight bullets, and a group of mismatched cases, I found that I can not shoot the difference.
Some can, obviously. I can not. If one can shoot this tight, than case segregation will shave tenths off of an average. Volume weighing being most exacting, and case weight being expeditious. And none of this will work if the rifle is not up to the task, as we are now venturing onto the border of custom territory.
Interestingly, the mismatched cases shot the best. I then understood the phrase "shoot the difference". I now save the advanced techniques for when I am able to. This gives me more time to practice, enabling me to reach the advanced techniques sooner.
I still separate bullets and cases, just while they are dry. And to the original question, I group them by three tenths. I still weigh powder on a Gempro and measure each bullet to the ogive. These things bring me confidence. Perhaps volume measurement will do the same for others.