Get rid of the microphonic tube. Something has become loose inside. If it shorts to another electrode, the whole amp could smoke.
> the glass envelope had become detached from the plastic/bakelite (?) valve base, and were basically free to 'flop about'...I didn't think that was something worth fixing with glue.
That's how it was put together: a brown phenolic/plaster "glue". Yeah, rough handling and overheating will break it. The tubes will work. You can glue them, though it will have to be a very high temperature glue. And unless you want to de-solder the base, glue, and re-solder (a royal pain), it has to be super-thin to seep into the gap. You could try cramming thin plaster in. You could try globbing the outside with high-temperature epoxy.
> one side of the waveform started flattening out at only 1 v p-p. The clipping was extremely unbalanced
Ah, that is a handy test. Put tubes in two at a time, no speaker load. Bring level up just to visible flattening. Most pairs should read symmetrical, and most should give similar results (maybe not 2V, but all the same). When you find an unbalanced pair, swap in one of the known-good tubes; if that does not fix it swap the other tube.
> balance of the phase splitter yet
If the resistors are within spec, it is good enough for rock-n-roll.
535V on the screen. A good Class A bias for 6L6 at that G2 voltage is -40V on G1. That seems to put you about 150% of plate dissipation rating, so no go. Class AB push-pull is less fussy, but -50V to -55V seems reasonable to me. Gives ~~75mA. But even that puts you at ~103% of plate dissipation rating. And the traditional rules of thumb (~70% of Pd) leads to 300 watts of heat at idle, plus 68 wats in the heaters, a stunning heat load. How do you want your steak? Very well done I hope.
No, the only way to pack 400 watts of audio into a stage amp (instead of an 8-foot rack) is to bias way down near cutoff. This does give some crossover distortion, but never as bad as a transistor amp.
OK, but how do you hold tubes near cutoff, when tubes vary a lot and especially near cutoff? Mesa's answer was to pick a typical voltage, and then select tubes to give a reasonable current at that voltage. Maybe 20mA-30mA. Considering the diversity of 6/side, maybe 15mA-45mA, though being careful not to end up with all tubes running 45mA.
I don't think "matching" is too important. I do think you need to keep all tubes near the 20mA-30mA mark, and keep the push and pull sides within 100mA of each other. (100mA would be a lot of unbalance in a normal amp, but this beast passes 2 Amp peaks.)
I would certainly consider a fan. Or more fan.