cheuschober

Thanks @PaulKuiper -- that definitely had me along the right track and was consistent with what I was thinking. After about 50-ish calibration prints I think what was happening was a confluence of multiple factors.

The biggest benefit came from upping the build volume temp from the default of 35C to 50C and changing fan speed from 15% to 65%. That had the most dramatic effect on what I suspect was warping due to the top/bottom having so much more plastic than the body and the forces related to that pulling inward.  Moving more air onto the freshly melted plastic and keeping a generally warmer chamber had a dramatic, but not complete, effect.

I then *also* had to dial down the skin flow by a couple % as you suggested to get a clean corner.