(Sorry for the really long post but figured this might help someone troubleshooting these boards down the road. Again - I'm not an expert; just learning as I go, but man I'm starting to get it!)
I just modded an older standard SI board I had for the same strapping, and it also worked with the 2716 and S2 to ground.
I also verified that the black box is a ram test, and indeed it works as I detailed earlier. With the board sitting on the bench face up and the daughter board to the left (And the ram towards the back), the farther back row is H to A (left to right), and the other row closer to you is 8 to 1 (left to right). It does round out numbers like 8 so if it looks like an 8, it could be an eight or a 6, but not E. Bad rams can take out columns so the number or letter maybe missing a piece. For a quick test try piggybacking the ram on top of the suspect ram (making sure the pins make a tight fit).
Each column of ram from left to right runs off each data bus from D7 to D0 (left to right) off of the 74LS174 at D7. Each data bus is connected to each column through the 330 ohm resistors above each column to pin 6 of each pair/column (A column would be 8 and H, 7 and G, etc).
I just fixed this board (woohoo!!) - once I replaced socket H and the cpu socket, put in a fresh cpu, it came up with the inverted lines like it should. Modded the strapping (bridges were actual traces...made tiny cuts that could be reversed if need be) and the rom started working and came up with the black box and most of the number 8. Wasn't sure so I replaced E first, which obviously didn't work. Replaced 8 and still got same thing, although the missing row from the number 8 would fill in when grounding pin 7 of 8 or H, which practically verifies something in that pair is bad. Replaced H, still no dice.
Did some comparisons with the logic probe and found pin 6 (Data bus D7) was low when all others were pulsing. Traced that back to pin 14 at D7. Replaced D7 - still no dice. Finally looked at the traces again, and found out they continue from pin 14. Followed it and found a solder bridge on a capacitor. Fixed that, and board came up to the test rom! Woohoo again!
Board was randomly resetting because I had it sitting on its side with no brackets on the daughterboard. Turned it upright and re-seated the daughterboard and 14 pin plug, and seems rock solid now.
Also while going through a few daughterboards I found the 14 pin plug can get pretty dirty and make a bad connection, causing resets. That 14 pin plug is responsible for the reset line as well as the 18v for sound.
Last but not least, the diodes responsible for the 18v on the power supply (thank you Elutz it finally sank in) which Elutz said were compatible with a 1n4007 - well I found Radio Shack's multi-pack of diodes had 4-5 of them in it, so I was able to fix both power supply boards I had and got sound again.