This is all subjective and so to many may be only slightly less than completely useless information but I want to give back whatever I can to the forum, and so far this is it.
I put the 2N7000's in the cathodes of the 5842's in the HF amp (vt-25 out). The sound is noticeably different with what at first sounds like a lot more clarity. The thing is, that rather than be a removal of something that was blocking or obscuring, this sense of clarity seems to be something of an artifact in itself. I really like the clarity, but but there is a certain loss of 'connection' - ie. me here and the music over there. Looked at another way, the musical topography is somehow monoplanar - Sorry, I know this may be meaningless subjective gobbledygook to everybody but it's the way I experience it and without the measuring instruments to put it into another language it's the only factual data I can give . In short the result is that I end up wanting to shut everything off after about 40 minutes with the feeling, "OK, I've had enough.". (One reminder that this still leaves open the question of whether it's the MOSFET's or the radical change in Ip that are causing the qualitative change.)
Next is to put the bypass caps I was using for the resistor bias across the MOSFETs to see how that works - then repeat in a couple of days with IR diodes.This will put the Ip back to where it was with the 75 Ohm Rk.
I also tried 25V Zener + 100 Ohm series resistor on the VT-25 cathodes. Sounded much better bypassed, but there's a certain understated stridency in the resulting sound so went back to a plain resistor, also bypassed.
Again, no technical info here but I hope it makes sense to somebody.