I had a partial case head separation using a 1.8oz buffer

[grayfox]

This catastrophic failure happened with new Lapua brass on his first firing of that batch, #4 of 50 hand-loaded cartridges.

That brings me to the question, can we see the heads of the 3 other fired pieces of brass if we haven’t seen them already?

I believe he said #41 out of 50...
But your point is well taken, what about the 3-4 preceding rounds? Also, we don't know how many fired, what rest time in between groups or individual rounds, maybe OP can fill us in.
At any rate, preceding rounds might give some insight.

Again, as to Lapua, in the AccurateShooter article it states that the older, muntz metal, brown box Lapua was reported to be very long-lasting... one cannot infer any one brass is better or worse (this from several metallurigists he queried) so all together, in my opinion any conclusions about metal or strength or brand, are speculative at best. The only physical thing we know is that it is thinner in design at the bottom curvature between wall and base.
Headspace, which develops into space at the back of the round after metal has expanded forward to fill the chamber volume (if I have that right), non-supporting cone-depth/shape/design, other tolerances combined- could translate to a too-long exposure of metal to hot pressurized gas, with possibly the bolt pulling away "early" due to unlocking sequence of a 12"/carbine barrel (ejecting at a higher pressure than a rlgs for ex)... at some point the combination got to the point where one case failed.

I would think that 2 or 3 of the preceding cases would also show some kind of beginning fail/incipience or imminent issue... if it were a purely mechanical/repeatable scenario.

Mention made of the millions of lapua rounds, so we also need to admit that these are the only reported instances. How many were unreported? How many unreported from other vendors'? Without some kind of methodical study even on this combo, it would be premature to point the finger at any single component, or even any single sequence of events. It is troubling for sure, but on balance, I don't see any generalizations honestly coming from what we know at this point.