Had some more time to work on this yesterday.
The Power section powers up! And nothing blows up!
The MOSFET family I'm using is the IRFBG30
I picked it out of the bin because all of the max values are acceptable, but there are some issues I need to keep in mind.
Foremost among them is that it takes 10v on Vgs to fully turn on. Max 20v on that junction. As it stands, I have a clean 5v rail and a dirty 80v rail. Eventually I may need a 12v rail for driving motors, but I'd like to put that off as much as possible.
The 5v section is isolated from this portion entirely, so I'll try using an optoisolator from my micro to feed 80v to a resistor divider, putting ~15v on the gate. How well this works will depend heavily on how well I can filter out the noise from arcing.
The transistor's max pulsed current is 12A, which is great (I plan on using 4A), but while I am technically doing pulses, it will be repetitive and as high duty cycle as my mechanical system will allow. These are rated for 2A continuous at high temperatures, so I may have to put to a couple in parallel.
If I come across a transistor that is slightly faster, lower Rds (this is 5 Ohm - too high!), lower max voltage but higher current, I would be pretty happy. Modern technologies are pretty great, I'm sure there are lots of options out there. I'm using components scavenged almost entirely from my hackspace's parts bins, however, so I am limited by what I can find.
That is made more difficult by the hackspace being in a state of flux at the moment after having our lease unexpectedly terminate recently.