Mike
I have written about why I built it and my many failures along the way and other stuff at: https://www.thetubecomputer.com/
All the tecky stuff is on the second page along with the new instruction set.
With vacuum tubes they warm up, start to work and then drift. Each tube drifts and so the whole system drifts. It takes quite a while to stabilise. I have set up all the tubes so that they hopefully become active NOR gates after a few minutes. This is when the tweeking begins. After each switch on a few tubes need to be adjusted to work as NOR gates and that's why there is so much monitoring
I am currently trying to improve stability of the entire system by redesigning the clock system and adding it to the bottom of the computer. Tests have gone well, and I hope to soon have a much more stable system requiring far less tweeking!
Check out the IBM 704 or 709 vacuum tube computers at the Computer History Museum in Mountain View, CA. They have a clam shell design. I have been inside them. On the outside there is a "wall" of vacuum tube modules behind clear, plastic sheets. On the inside there is yellow wire connecting the modules. I think the tubes are on the outside so one can see any that have failed.