Bass
As to the Red Dot test; the following also applies to it. If you remember the RPM threshold applies to otherwise accurate cast bullet rifles loads using medium and slow burning that are pushing higher RPM because of twist rate and velocity. We "can make" bullets/loads to be inaccurate, particularly with fast powders. By causing a lot of deformation and unbalancing through rapid acceleration the bullets, at the low twist rate of a 45 cal barrel, never attain complete rotational stability before the centrifugal force of the RPM overcame it. This occurs even at low RPM as you discovered.
RPM are only linear given the same twist with increasing or decreasing velocity. However, centrifugal force also increases as the RPMs increase. I'm not sure but perhaps the increased centrifugal force is not linear as the RPM increases (anyone know? I don’t have time to research it today.). I expect it is not as accuracy is not linear even with the same twist, i. e. the greater the RPM the inaccuracy is proportionally greater.
We do know that proportional accuracy is not linear with the same twist or when compared to other twists. Not only do my tests adequately proved that but I’m sure you and many others have also observed it in your own tests. I thoroughly expect the additional comparison tests of the 3 twists will show the same nonlinear inaccuracy.
Larry Gibson