The project team is very aware of this issue and the ill will it has caused elsewhere. As an issue here, it will become much less likely as more projects are brought on board.

However, it looks possible for a computer to receive identical WUs a moment apart, since my older computer did more than once. I mean, same number, same size, I checked their content was the same with diff.
I noticed that right away in UDMon and manually deleted the duplicate WUs.

I know UDMon can corrupt WUs, but it will not corrupt slots that way. Indeed, I noticed some time ago it does corrupt WUs on this computer, whose HD is slow: all no-valid-reason aborts I noticed (I posted about them more than once on this board) occurred in a 10-13 seconds time frame after a UDMon backup, and after restoring the folder from an older backup, the WU can reach its completion. This makes me think of a race condition, with a lock released too late, which makes the UD agent abort everything.
I never spent time reporting this bug to the UDMon programmer, because I was too busy, but I guess I should do it now that I'm more or less on holiday.
Therefore, I disabled backups on this computer, even more so now with the enormous intermediate files of FightAIDS@Home.

> I'm glad to see that you're on top of things enough to recognize when it's causing such an abnormal behavior and remedy the problem.
I'm not always good at that, actually, but as a CS student, I'd better be able to find out, and if possible solve, causes of abnormal behaviour ;-)