I'm not at home right now, but I suspect that something went wrong after your computer uploaded the result. This prevented it from reporting the task as complete. At any other BOINC project that would have been the end, and the task would have stayed as an error and you would not have gotten credit.

We have a special process looking for orphaned results such as yours. It fixes them so that they can be validated. That's why it changed from "error" to "valid".

The system did send out a replacement task, but that happened immediately after your task was initially marked as an error and can't easily be prevented.

____________
My lucky number is 75898⁵²⁴²⁸⁸+1

That's exactly what happened. Given that the residue has 18446744073709551616 possible different values we trust that if two results match then we can safely ignore the error flag.

You hit famous "finish file present too long" Boinc bug. Look like that for unknown reason, Boinc tried to start (most probably restart same one) task while another instance was still running in same directory. Meantime, first copy successfully finished and uploaded valid result file, while Boinc was confused that something is still running and reported an error status to server.

I see that LLR process was restarted many times. Do you have a option like "Pause computation when computer is active" turned on? Boinc has a design flaw when termination of task at specific moment will trigger race condition - Boinc will not notice that task has already finished, restart it from beginning and finally abort with "finish file present too long" error.