on this:Your read of using the driven side plate resistor to adjust the gain (when output is taken off the non-driven side) is brilliant; I hadn't seen that.

I had not seen it either. Seems like all sorts of possibilities open to play with that one. Two completely different types. 12GN7 pentode on the input side and a triode strapped version getting its cathode driven for one extreme. Play with the pentode's plate load and pay particular attention to how it is loaded and I'll bet it could do some remarkable stuff.

Two matched vlaves might be a better option. The plate load thing can be adjusted at both sides depending on the gain required. Low plate load onthe valve with its grid taking signal would offer an inverting low gain side. Or put the lower plate load on the grounded grid side. The output should be taken from the 'side' with its grid grounded for appearance's sake at least.

Just look at all the math applied to describing the balance in a Schmitt inverter and use it to quantify the imbalance that is required. Small common cathode loads( *NOT* CCS-like ) could also be applied to tuning the circuit performance.

Going to have to play with this stuff sooner or later, no doubt about that...
regards,
Douglas