I had a hunch, not sure where it came from, that the MCP120 reset controller was to blame. I set up my probes and then did a quick continuity test to make sure they weren't touching each other... BEEEEEP.
Okay, so I connected the ground to the serial card (which had an easy to connect pin) and BEEEEEEP.
Some more debugging, including grabbing an unpopulated board, and it it looks like RESET is somehow tied to GND. That's a problem, because it means the board will never let the CPU start. I haven't figured out how this is happening yet.
For reference, here's the typical applications section of the MCP120 datasheet (I have the MCP120-485HI/TO):
And here's the relevant part of the schematic:
J1 is a reset button connector so I could break it out to the case. On the board, it's an unjumpered header.
I still need to figure out why this is happening and how to fix it; I don't have money to buy a new board until December, probably.