I’m trying to design a PCB to last, and there’s very little information publicly available. There’s the obvious - no electrolytic caps, no pure tin, sure - but beyond that, what would you do?
Also think this question would be good for retro computing groups. They can tell you what their common failures are so you could avoid them.
Meaning: you are missing a lot of variables in your game that will mess with your plans.
So spend some times and do requirements engineering on an abstract level. Then select the techniques to solve them.