A few of my own thoughts.

Kaase's P51 heads are excellent and not just because of the flow. They don't flow anywhere near 400 unless you test @ 10" and convert up. However the flow @ .400" is remarkable--I believe 325-330 cfm IIRC. It is the chamber and how it is positioned (which you pointed out) which belies its true flow capability. What is baffling about them is the massive stall they suffer from @ .550" and up. I've tested 3 different sets and they all exhibit the same trend. Is there something more to this?

Those curves you put up are all indicative of a large port volume AND the wrong header design. My 2013 466 Pro-comp Victor deal had the same trend until we added something like 24" of collector with a .5" reduction in diameter along the way. This knocked out most of that dip. However this is pretty much a Band-Aid to a high flow port that is too large for the application. Min CSA of these heads were exactly 3.125" as calculated by Solidworks.

Learning from this I used a 2.85" CSA on my 2014 EMC 439.9 ci (440) effort. These Victors were dowel shifted, angle milled, guides moved in X and Y and angle changed. Flow was 353 cf. with a 2.19" valve. Torque was impressive 535 tq @ 3000, off the brake zero dip was observed and a peak of 630 was observed. If I could do it again I would make the port even smaller. That engine was easily 100pts better than my 2013 effort. J.Rob