I can see what you mean.
However, I'd imagined that at the moment of firing the case would likely be shoved forward nearer (if not onto) the shoulder of the chamber by the striker anyway - at least, as far as the extractor will allow; and then the rapid rise in pressure would cause the case-walls to grip the chamber - thinnest bits of case-wall first, all the way back to the web - until finally the web/case-head (which doesn't expand/grip at low-enough pressure to prevent this) is pushed against the bolt-face resulting in the familiar thinning in front of the web.
Any road up, I think that annealing like to be a very distant second to setting the dies correctly when it comes to reducing the incidence of case-head separation.