In theory, that diode junction does nothing, but he claims a small audible difference.
Could it be something to do with the junction capacitance of the diodes?
There is the hypothetical consequence of such connections, and there is the audible in-the-field effect. Some effects which are heard ARE illusory, and the byroduct of human expectations. But some effects are real, only the byproduct of either rare, or special circumstances; this kind of pickup, with these guitar or pedal settings, and that kind of amp at that volume level. I won't dispute the musical relevance of such unusual combinations of factors, since many of the recorded sounds lodged in our collective memory as benchmark tones DO come from "one-off" circumstances. But the average player in the average playing situation, whether bedroom, bar, music store auditioning room, or basement band, is not going to notice.
Assuming all schematics posted around are accurate, the Timmy is distinguished by three things:
1) The use of a 2+2 diode complement, resulting in less clipping (or at least clipping more confined to initial pick attack), and more output level.
2) The variable "ground leg" of the clipping stage that alters bass response of the unit.
3) The use of a variable lowpass filterfor tone control, instead of the complementary treble-cut/treble-boost control of the TS-9/SD-1/similar.
The reason the Timmy is so appreciated by players is that these three features allow for one to push the amp a little harder and use the amp's overdriven sound in tandem with the pedal's, allow the bass content to be matched to the pickups and amp settings, and allow for a "smoother" sounding output with more dialable treble-cut.