My surmise has been that these problem loads experience uneven ignition delays in any revolver, regardless of barrel length, due to primers unseating the bullets to an irregular degree before the powder really gets burning, and that in that circumstance the bullet base location in the barrel at the moment the peak pressure is reached has a ±tolerance that, with slower igniting powders, includes a point about 2" down the tube.
I assumed the same, when I encountered the issues with Lil' Gun.
Because... I wasn't getting "fused powder" with light charges, but heavy charges.
I was working up a handload for .22 WMR, from something like 3.5 gr to 7.5 gr. From the starting charge to about 5.5 gr, it worked well, but had low muzzle velocity. But, once I increased the charge to anything greater than that, erratic burns became the norm and I stuck quite a few bullets in the bore (anywhere from 1/2" to 2" from the case mouth).
Based on the evidence I had available at the time, I decided it all came down to one basic factor:
Primers unseating the bullets and starting them in the throat/bore, causing the powder to 'fizzle' with the loss of pressure after ignition. This also resulted in the loss of chamber seal, if the bullet fully entered the bore.
The more powder I had in the case, the farther the bullet would be pushed into the bore by the primer. And... the farther the bullet got pushed, the greater the sudden loss of pressure was, resulting in the 'fizzle'.