I'm working on a Kiss LE, and the demon head seems to correctly pass a locked ball from the VUK up into the head, but then it immediately releases it rather than holding it in the lock, and no lock is awarded.
There is currently one bad opto in the head. But, it is the third (furthest toward the back) opto, and it's stuck in the "no ball" state. I.e., the receiver always thinks it sees a beam. I would expect this to cause the first two balls to lock normally, then the third one to go undetected.
But maybe the machine is trying to outsmart me: perhaps when it doesn't see a descending series of switch hits 3-2-1 when the ball passes, it assumes a bad opto, and disables ball locking altogether? Unfortunately, we don't have a spare opto on hand, so I can't test this theory. (And anyone with a working game would need to unscrew one of their receivers and shine a bright light into it to simulate the situation.) Seems like a better failure mode would be to release the ball but count a virtual lock.