I believe I have diagnosed the problem, but I do not know where to begin to fix it.

In test mode under Switch Test, display reads All Switches Open (with having removed the balls). All switches test correctly except those that are on Strobe Line 5. In test, if I hold a switch closed on Strobe 5, the display will read that all switches on the corresponding Return Line are closed EXCEPT for the actual switch I am closing.

Example:
Close switch 35, display reads "35". Close switch 55, display reads "05 15 25 35 45 65 75", but not 55.
Close switch 24, display reads "24". Close switch 54, display reads "04 14 24 34 44 64 74", but not 54.
Same for the Tilt, which is on Strobe 5. The Tilt is on Return 7 but test displays "07 17 27 37 47 67 77" and 57 (Tilt) is the only switch on that Return.

If I try to start a game, the same holds true: switch 5X will not register but rather will trigger switch 2X, 3X, 4X, 6X, 7X (and theoretically 0X and 1X, but there are no switches on those Strobes) .

I read the Dear Lloyd article post where someone was having trouble with an 80B matrix (Victory) and Lloyd said to turn the game on with A1J6 disconnected and then under switch test, the display should indicate if a strobe or return were faulty. When I tried this, the display still read All Switches Open.

..and the 80B adventure continues!

Not one to rest on my laurels, I went straight from here to pinwiki.

I thINK I've located the culprit to be a bad IC, Z12. Even though every leg read as it should across Z11-Z15 in diode test on my DMM, in continuity test, one leg of Z12 read short to the ground leg.

I'll stop by my local hack space (milwaukeemakerspace.org) tonight and see if they have any 7404N's..

Three cheers for Swemmer boards having all IC's socketed and easy to service!

Huzzah! I got the ICs I needed (and a few extra spares). Replaced Z12 and Z11. Everything back to operational again!
Idk why, but even though it appeared to me that only Z12 tested faulty, I had to replace both Z11 and Z12 to get it to work. Or maybe it could've been only Z11, but I changed 12 first to no avail and then changed 11. Either way, goes to show testing components in-circuit can lead to false readings.