Sorry, no photos. And I didn't do that one, factory lower but it was machine staked, squared off dents. I was not happy with the look of the threads afterwards, so I tossed the receiver extension, endplate, and castle nut, did it all again, and used a center punch to stake the hell out of it.
I did have a weird situation. It was so muddy I couldn't find a bit of ground that worked, so whacked it on a log used as a barrier, and it fell into the crotch instead, so the butt was sorta forced to an angle as I brought it down. Call it all my fault
Just using it as an extreme example, that you better do it right because even a really hard match day can stress the system. I did twist it by hand back, and ran the rest of the day with it hand tightened.
I have heard secondhand stories of 3-4 similar failures of properly assembled guns in the US Mil and one name brand police one, that got fallen on, dropped, etc. and loosened/rotated a bit. But not to stoppages.
OTOH, seen (myself) unstaked systems loosen so much the rcvr extension moves under load, and the buffer binds on recoil and the gun stops.
Also: do get good quality parts. Cheap castle nuts totally don't work as well.
Also also: I have a
vague recollection of someone (e.g. LMT?) who doesn't stake, insists that proper torque is enough, but has good parts, good process, and tools to torque it right. Anyone else heard anything like that?