which showed an appalling 'spread' of measurements.
Testing them to give the pentode curves I found that there were two clear groupings of curves, I suspect if you measured these at a single point it might look like a spread.
IMHO, wide spread of measurements can be handled either by:
1. Design that copes with that spread (though that can go against the tube dogma of simplicity)
2. They are cheap and available in numbers were you can select matched sets (jfets fall into this category).
I feel that as the availability of a lot of these things tends to zero many of the traditional choices and designs will become impractical. When I was working on the ref design I originally used half of the 12ay7 as the cathode follower that drove the g2 of the d3a, but I found that only the EH gold ones were quiet enough to use that way, and then I found EH were stopping making them (I bought a box of them direct from EH, just because I found they were quieter even though the final design didn't need them).
To join this to a current PITA, a lot of modern parts are getting hard or impossible to find. I have stopped making the turntable supply now because LM3886's are unobtainable for at least the next year or so (if not forever), and can't be arsed redesigning to use something else. I have just bought 2k of a mosfet that is end of life that I use for the two box power supply. May buy some more while I can but I will eventually retire so such is life.
Little known fact, coherent thought can destructively interfere with itself leaving no thought at all, that’s why I prefer incoherent thought.