Okay, let's see if that made a difference.
Interestingly, the system uptime for the latest crash shows:
so I'm guessing the computer was shutdown before the CPU was removed and then replaced such that the system state was saved to disk in a hibernate state rather than a full shutdown. So, it...