I managed to track down enough components to get my MiSTDeX card working fully, but not without a ton of bodges. Always check your schematics over and over again, trace through your address logic, and make sure you can actually get the right parts.
First problem I ran into was not having enough 74_574's on hand to make the address latching circuit for the VDP's memory. Turns out you can use a 74_573 in place of the 74_574 that buffers the VDP's data output into the SRAM data bus if you tie the 573's latch enable high. You could also use other bus drivers in a pinch, but the 573's pinout is close enough to make it a single pin bodge.
Next, I had ordered the only 10.738MHz crystal I could track down, but it was the wrong type (series vs parallel) for the TMS9918A. This necessitated building a 74_04 based clock generator circuit, and outputting opposing phases to the TMS9918A. It is close enough to get a picture, but the colorburst is just off enough to result in a black&white picture on the TV. I'll keep trying to track down the correct crystals, or a small variable capacitor to tune the frequency of my bodge.
This brings us to the address logic. Turns out I swapped A4 and A6 on the address logic for the PSG and VDP. This results in a horrible mess trying to figure out which jumper to use for a given address. And I then noticed that I had made several other mistakes in the PSG address section.
Now the good news, I have managed to get RomWBW to work with the TMS9918A, and the included chiptune player works on the AY-3-8910. I've also gotten J.B.Langston's demos working on my hardware.