In general, if you are setting out to do something, there is a very good chance that someone has already taken a trip down that same road before you. I recommend looking for their tracks, it often helps you avoid a landmine.
A little googling and I've come across a few pioneers. One of particular interest is a Masters Thesis by a fellow name John Bork that hits fairly square on what we want to accomplish.
This was also written up as an article for Linux Journal which is how I came across it.
The best part about finding this work is that my heretofore unfounded hopes for a successful outcome have been (a wee bit) validated. John's work had a different goal in mind, but it clearly demonstrates that with a little persistence, we will be able to build a new (dare I say better?) controller and leverage the existing I/O driver board of the machine. He also validates that we will be able to successfully drive the Lamp and Switch Matrices.
Don't worry, there is still bountiful opportunity for spectacular failures along the way. This may be a fine time for everyone to pop in a DVD of Meet the Robinsons.
Because like them, no matter what, we will "Keep Moving Forward."