First off, Rick, thanks for the circuit. It looks interesting, to say the least.
Next, skipping over the potential Jack/Rick's GF controversy, I wanted to inquire as to the theory behind the BS170 Drain biasing of the SH-SHO. I mocked up a quick SPICE model of the circuit, wanting to see what the voltage changes were at the Drain of the BS170. (Unfortunately, the crappy SPICE tool I used didn't have BS170 or 2N5457 in its library, so I modeled with IRF250s.)
Using the DMM tool, I measured DC voltage at the Drain of the BS170. At lowest-gain settings on the 5K pot (95%), voltages were about 8.2V w/HOT side and 7.6V w/SUPERHEATED. At mid-travel (50%), voltages were about 8.6V w/HOT and 8.4V w/SUPERHEATED. At highest-gain settings(5%), voltages were about 8.7V w/HOT side and 8.6V w/SUPERHEATED.
I was curious as to the design theory that contributes to the increased gain from the SUPERHEATED option. It appears to reduce BS170 Drain voltage at low-gain settings, but almost-mirror the voltages from the HOT option at higher-gain settings. I myself do not have an explanation for why this causes the gain increase from HOT to SUPERHEATED, so I was hoping you might fill in the gaps in my understanding. (Pretty please, with single-malt scotch and Mullard tubes sprinkled on top.

)
Thanks in advance.