No idea based on your comments, but stream of consciousness and food for thought....
If it is important, stop running that batch if you don't have time at that moment to slow down and troubleshoot.
Any time you encounter a function issue, save that brass and isolate it. Also, keep samples of "good".
You will want to do a very close inspection on it, versus ones that function normally.
The drawings are on the web, and it doesn't require sophisticated lab facilities to run a micrometer and caliper.
It pays to have a chronograph handy to catch examples that are out of family in terms of fast or slow and determine if the failures correlate to those. It can be hard to catch unless it happens reliably.
(Now that chronos are smaller and work from your side of the line, it is trivial to set one up and take data. I don't always carry one, but I do when working with anything new, pistol or recipe.)
Failing to find issues with the ammo in terms of dimensions or velocity, that leaves us with a focus on the pistols.
I agree the odds are bad that two of them with a clean history on factory ammo are not likely to blame this time, so I would focus on the loads.
There used to be an old saying in the troubleshooting business, that the simplest solution is usually the explanation.
If you have not already done so, I would take the rest of the ammo batch and drop it all into a min-chamber cartridge gage to make sure it passes minimum chamber dimensions. If you don't own one, use the barrels.
Take a close exam of the rims on everything you have left of that batch, loads or empties.
If you have a borescope, check the barrels up front to make sure they are clean and no issues. Check the extractor function, bolt face, and firing pin hole. On the next test run, isolate and save samples of "good" and "bad" brass.
Good Luck and in for the range report.
