• WARNING: Tube/Valve amplifiers use potentially LETHAL HIGH VOLTAGES.
    Building, troubleshooting and testing of these amplifiers should only be
    performed by someone who is thoroughly familiar with
    the safety precautions around high voltages.

Puzzle: Order of magnitude drop in THD+N around 10 kHz

So, following on my journey, ran a test connecting the probe to the wiper output of one channel of the pot, and with the power cord disconnected. (no amplification test of course)
It removed the distortion in large part.

So, not the pot, but interference from ground.

I disconnected star ground from earth ground and ran another test with power connected and volume knob at noon.
In that configuration, distortion is down to 2% from much higher levels before.
BUT, I get very audible hum when no music is playing. 😒

Also, even though distortion is much less in these tests between 20 Hz and 10 kHz, I still observe a sharp drop in distortion past 10 kHz.

That amp rolls off the highs...20 kHz is at -3 dBr and it drops even faster after that, so that maybe why the AD2 cannot measure distortion correctly after 10 kHz.

My next test will connect the amp to a variac pushing out the 110V the PT really expects...will see what that gives.
 
Your tests confirm that your setup detects distortion below 10kHz, but that's it: it could under-evaluate THD > 10kHz, you wouldn't notice, and that is probably what happens.
To confirm that hypothesis, you can test an aperiodic THD generator, like a softened diode clipper: if you see a drop beyond 10k, you know that your measurement system has a problem
 
Your tests confirm that your setup detects distortion below 10kHz, but that's it: it could under-evaluate THD > 10kHz, you wouldn't notice, and that is probably what happens.
To confirm that hypothesis, you can test an aperiodic THD generator, like a softened diode clipper: if you see a drop beyond 10k, you know that your measurement system has a problem

I have come to accept the amp probably doesn’t have enough bandwidth to measure THD past 10 kHz.
 
After retracing my steps and checking everything, I think it’s just a matter of not exceeding a certain input signal level.

In particular, it seems this headphone amp is not ready to accept 2 Vrms that modern DACs put, while it’s volume knob is at the noon position.

But at 10AM, distortion is below 1.5% and the amp sounds quite good, provided that high impedance headphones are used.

As far as measuring THD+N past 10 kHz, it might not be feasible with this particular amp, but it works fine with several other solid state amps.

Thanks to all of you for your help!