banzaibird, for getting estimates of performance, I use both the Powley Computer and QuickLOAD. While I have a set of slide rule Powley's, I use instead a version I wrote from the equations presented in the NRA Handloading book, with corrections made from other sources. The equations in the slide rules are different, and probably represent an earlier attempy by Powley. The slide rules are not as accurate for predicting charges and can be downright dangerous.

Not having a strain gauge rig, I had to compare Powley results to published pressure data, from Hodgdon, Lyman, IMR, etc. I find exactly what you did, that Powley's pressure estimations are off as you near 50,000 CUP and are not reliable below 30,000 CUP as well.

For case capacity estimation in a wildcat, I use a calculator based on an idea Ken Howell gives in his big book on cartridge conversions. It's linked from the Powley Computer just cited.

QuickLoad is more reliable than Powley, but its author acknowledges accuracy falls off at lower pressures. Frankly, I think the Powley does a better job in this range.

As for powder selection, no software tool seems to be able to distinguish between similar powders. By similar powders, I mean those near each other in any of the published tables of "burning rates." I think only with a strain gauge can one sort out minor differences. Unfortuantely, since minor difference are what separate a top load from the average loads, I believe no software can safely predict the powder for top performance. I take what the software says and drop about 5% in fps.

Powder selection can, it seems to me, be aided by studying load book data for similar cartridges. I elaborate on this on another link from my Powley Computer. It involves computing something like bullet SD, but for case capacity instead.

Karl