I have no extensive documentation. Heck I've never mentioned this to anyone that was not in the room at the time, but in 1996 - I did this. With a smaller disk - an 8" disk mounted high in old tower PC (newer than the disk), and some x86 assembler.

This feels like physics to me - I'm not sure why people need extensive documentation to believe that throwing a drive head back and forth would cause a machine to vibrate (in fact even at just 8" those large drive heads make a machine jump and stutter) and if one was lucky do so consistently enough to walk the machine forward.

Considering that only recently the first-wave of RTX 3xxx cards had a flaw where uncapped FPS on videogame menu screens caused them to spike current outside of the intended design envelope and completely brick the devices with unfixable damage, I think it's pretty safe to assume there's a long history of "user doing something programmer didn't expect ends very badly".