That makes a little bit of sense. IPDB only has 1 page of the switch matrix, here's what I was using from another pinside thread:
https://pinside.com/pinball/forum/topic/dd-switch-matrix-need-help#post-2595613
I think the diode thing messed me up more than it should have, in combination with the wire colors being different after the coin door connector. Here's what I've had going on.
I've been having an intermittent problem with multiball being awarded when it should not (sending the ball around on the skill shot after locking the ball or on initial plunge will sometimes award multiball when it should not). I recently completely rebuilt the skill shot 3 (top) gate and the Level (final gate on the left ramp). Today I unplugged the connectors and checked every diode, verified the involved switches worked properly, made sure all switches showed up as they should in test. This is where things started to got weird.
When testing to see if anything helped I started having what appear to be switch matrix issues but with little rhyme or reason. Most switches on CJ4-14 cause the action of the switch to the right of them in the matrix (coins and return left caused pop bumpers, return right caused left sling), but the rebound switch would cause a slam tilt which is in the same column. This would happen in game but in test mode the switches and coil tests worked perfectly. I removed the playfield connector and bridged the rebound switch pins and it did not cause the slam tilt. The problem was intermittent so these two items could possibly be red herrings but I have never gotten a switch error to show up in test mode, activating as many simultaneous as I could and doing my best to recreate game conditions. This could be due to the crappy nature of the 6803 switch test not showing a momentary closure of a problem switch.
After this I tried replacing the two socketed 6801 chips and replaced the switch IDC connectors with crimp and stuff with no changes. Without changing anything else that I know of the switch matrix problem seems to have gone away with the multiball problem remaining. During this process I tried to look at the involved columns and check for lugs and leafs being contacted, and loose diodes / diode legs contacting lugs etc. I'm not sure where to go from here but I'm getting ready to just replace every diode in the MFer, starting with the rebound switch column. I'm looking at the matrix trying to find a potentially bad diode using the box method but I can't find any commonality with normally closed (or closed during gameplay) switches that makes any sense. I'd previously snipped a leg off of most of the capacitors on the game trying to diagnose the aforementioned multiball issue.
Is it possible (or probable) that this is a logic issue, not a switch issue? The windows in the game ROMs are exposed but from what I read this isn't usually a problem.