Concerning the video above, unfortunately it’s probably difficult to write a mathematical representation of a tube amp’s transfer function so it’s difficult to use mathematical tools to determine gain and phase margin. It is however rather straightforward to test it with the actual amp on the bench, and you can determine rather quickly through a Bode plot what the actual amp stability characteristics are.
That being said, I don’t have tools to make Bode plots but I still know my amps are stable. You do this by trial and error testing on the bench, knowing for a transformer coupled tube amp, phase shifts approaching 180 degrees occur at the frequency extremes—down in the subsonic region and in the supersonic region. You then apply measures to either reduce phase shift, or reduce gain to less than unity, or both, at those frequencies, to force the amp into stability.
Control theory at large is a complicated subject and people get PhD’s in this stuff. For tube amps it’s rather straight forward once you know a few principles.
. I've built a lot of basic, no-feedback SE designs before, but when I tried to figure out a PP feedback amp this really helped, even if I had to have a gadget do the work for me. ;-) The fact that he uses a basic Williamson as an example, the very amp I was trying to build, was a bonus.
. After 3 hours the PT is warm, but not uncomfortable to touch - infrared scanner says it's 122F . Absolutely no bad behavior that I can see at this point. It is very sensitive and will buzz if the input is open. With a shorting plug or connected to a source, it is very quiet. I hear nothing in the horn and a very slight buzz in the woofer. Once I get the scope working I can hunt that down. 
