Can you explain the theory & results of:
* rather large filter caps for nodes A & B
Most of my life I have been dealing with large guitar amplifiers. I learned early on that you need a lot of filter for the power tubes if you're going to be overdriving them. Otherwise you get hum while overdriving that is particularly annoying on leads. I was putting 470uF for the first node on some of those amps. I don't have much experience with overdriving 6V6's, but their power demands are much lower and I decided to try 100uF to see what happens.
* reduced & "nearly regulated" screen voltage
There has been a lot of discussion about grossly exceeding the published voltage limits on a 6V6 and I have actually had some trouble with 6V6's at 450V. Some of the bigger bottles can take higher plate voltages as long as the screen voltage is reduced, so what about 6V6's? I'm trying it out.
That 22772 PT is rated a little low for this amp as well as for stock Princetons. Moving all of the lines down on the graph with reduced G2 voltage in order to lower the current consumption seemed like a good idea.
* the dual 500k pot
I only have three holes in the faceplate, so the volume control needs to work with both Input 1 and Input 2. I always put a volume control after the first stage because distortion created by the interaction of the first and second stages cannot be eliminated with a volume control after the second stage. If I just used a single 500K volume control after the second stage, I would lose some versatility.