Agree that bridged pins are most likely, but you can get some odd behavior with any break in switch row continuity. I spent a long time with this board a while back and while troubleshooting found that I could place a partially-fried LM339 in different sockets on the board and the problem wouldn't necessarily move to the optos associated with the socket it was in at any given time. I.e., I could put the bad chip in socket one and see different behavior on the optos wired to socket three.
Long story short, might as well check everything while you've got it out.