#1). The detonation thing was with extremely slow powders. 4831 surplus in particular. Mr. Ackley said he could never "do it" (in his two volume set of books available from eabco.com, sinclairintl.com, ebay, etc.) but his best guess, the primer was packing the powder in the forward of the case without complete ignition and then as the powder plug burned and pressure built against the plug/bullet, things would let go... 4320/4064 powders, stick or ball, faster are not supposed to be subject to this problem. I have seen specualtion that if you were silly enough to use a pistol primer in a rifle case... MAYBE... but I repeat... silly...
#2). Lyman manuals for many years have recommended cutting loads with 4895 down to 3/5's book max. No loss of accuracy, just velocity. Hodgdon takes questions like this by email I believe they usually recommend 4895.
I would not worry too much about cutting R7 or 3031 loads some. Dupont/IMR had a data manual (free) that listed loads with all IMR powders for each cartridge. 4831 loads in the .22 Hornet are pretty silly. (or the .223...) And the 4227 loads for .458 Mag were tiny...
And you can achieve these reduce velocities and kick by using a very slow powder. '06 case full of 4831 of 7828 or 870 under a 150 grain bullet is not going to make alot of velocity. THIS IS DIRTY. The unburned powder fertilizes the forest floor. But it is a possibility. Since you have a case full, detonation is not a concern.
Or you can use alot faster powders and lead bullets. Lots of cast bullet loads out there and now with cowboy action shooting... luck.