Working with JFET tolerances

[floobydust]

I think it depends on the configuration i.e. a source-follower changing the source resistor does not move Q-point (it's dictated by V_GS up the I_DSS limit) so I bias the gate up and a trimpot makes it easy.

Single discrete JFETS have fallen out of favour because their variances are too wide in the majority of configurations, in a production environment . If I use any in a hobby project, they are hand selected.
The only common JFET use today I know of is for electret condensor mikes, typically using the 2SK596 which comes binned in four I_DSS classes around 2:1 (and has internal gate resistor). Otherwise it's 10:1 variation with older JFETs. Newer JFE150 is ~2:1 though. I did not look at V_GS(off) variations, they are important too.
I think the electret mic buffer works (in a production environment) because a source-follower with low signal the Q-point isn't important, it will never clip with the mV output.

For guitarists that like to overload the guitar amplifier's front end, TL072 CM voltage can cause phase-inversion (9.3.2) adds a harsh distortion as well as its non-linear gain in some configurations. It's low cost but not great IMHO, usually gets nailed with ESD as well. Bring on the 12AX7 or a JFET lol.
I thought op-amps with JFET front-ends are usually laser trimmed, like OPA637 which adds a lot to cost. But no trimming can I see with the TL072 die-shot.