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

[LRRPF52]

I am aligning to your thinking. Basically I was evaluating the difference in how much a round, with an out of spec deep bolt and with excessive headspace, would add to the unsupported side wall area. 0.138" was your out of spec condition, and 0.133" was your in spec condition. This means at least 0.005" of case depth is being exposed to unsupported wall area, in addition to what is already exposed by the feed cone. I hope this makes sense now.

And then you could even add the headspace excess to it.

What I'm trying to do here is get a rough quantification of the importance of excessive headspace being the issue.

I’m tracking now. That makes perfect sense. You were basically establishing a window for tolerance deviation.