The setback occurs in all rifle cases. Hatcher measured that a .30-06 case shoulder could be set back by up to 0.006" just by vigorous fast chambering in the Enfield 1917. That would mean the inertia from being thrown forward into the chamber was doing it. Mind you, back then the good match shooters could really work a bolt rapidly. I saw Jeff Cooper demonstrate it on three different visits to the sconce. "Putta-tap". Opened, ejected and closed on the next round.
The key difference in the lead loads is pressure and how quickly the bullet gets moving down the tube. Roughly speaking, if you load to a peak pressure lower than around 30
kpsi, the case never sticks to the chamber wall and winds up backing up to the breech rather than fire-forming to the chamber and stretching at the usual pressure ring location to reach the bolt face. This leaves it small. A heavy bullet jammed in the lands does two things. The jam prevents the firing pin and primer from forcing the shoulder forward, so that what pressure there is, is applied to the forward shoulder, which is the thinnest, and from annealing, often the softest part. The big bullet gets moving slowly so there is a lot of time for the peak pressure to be reached before the bullet has moved much and for the brass to relax a bit into that new shape and be more inclined to take a set and keep it.
I ran into this with jacketed bullets and Trail Boss. I had recommended the powder for fire-forming, but one of the members couldn't get it to work. With some experimentation, I found the bullets had to be heavy, as the ones I'd been using up were. Having them jammed into the lands would raise the peak about another 20% in most instances. So some combination would be fine, too.