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

[StoneHendge]

That's the first point of my failure model! I have not neglected this...

"...it's a tolerance stacking issue...the barrel feed cone was too long,..."

Edit: To me, in hindsight, and because Bill Alexander expressed it as a necessary design change from the M16 feed cone, I don't understand why he didn't make that a specification. At least create a maximum depth spec. This is in hindsight, but I think it needs a spec.

It's got nothing to do with the spec. The manufacturer likely effed up making the barrel. We've seen it here countless times on here barrel issues which were likely due to reaming too shallow or too deep. Easy enough to do with whatever tool they do the feed cone with too.