Ok, thanks.
Now I'm waiting for another backup with DSOFlash to end, just to be sure before starting anything.
I'll take a new look at the ribbon cable, but I don't think that is the case, as the pushbutton signals travel through the same path as those of the encoders (see attachments).
The yellow signals go to the ribbon cable, the green ones correspond to pushbuttons (including those in the encoders, at Scan_Data_5), and the red ones to the rotary part of the encoders. If there's a bad connection, it should be between the encoders' diodes and U2, and to see that I have to take apart the display, a bit more complicated than accessing the main board, but I'll do it when the backup ends.
BTW, I've had problems with the FEL driver in this PC, followed all the instructions but DSOFlash gave an error. I've done the same on another PC and it's going OK.
EDIT: Backup finisher, took apart the display and reseated the ribbon cable. Also, checked with the multimeter continuity from all the diodes' anodes to pins 1 and 2 of U2. All correct. I also took the opportunity to clean the board, which had a lot of flux residues (I thought it could have been the culprit, if these residues had become conductive with humidity, for example). But no luck. I'll try platform tools