I'd love to be able to see what the IC on those daughter boards is. My best guess is that they are monitoring LED activity to detect a lockup, then resetting the downstream LED controllers.

If this is their "solution", I'd really like to at least see that daughter board merged into the main LED board to eliminate the additional connector and the jumper wire that you can see from between the two boards.

So, from what you guys are saying, and looking at B9's video, it looks like his #19 is still working (kinda tough to tell), but W1 locked up. W1 would be the one that needed replacing in this case?

Ok, I watched the video another 10 times or so, and decided #19 is locked as well. Either #15 or #16? Imagine trying to do this without a video!

So, try swapping the new board at #23 with the old one at #16 to see if the next lock up takes out less of the LEDs?

That's pretty cool. I'm assuming that since it's not quite the same as the conditions inside the game, it can't really be used to detect boards that will fail or lock up in the game, though, correct?