Tonebender MK II transistor gain questions

[MarkL]

I've made a few of these, and have gone with what seems like the traditional values of ~70 for Q1, 70-80 for Q2, and 100-120 for Q3.  I've used Tungsram AC125s in these and generally, they've sounded really darned good with a few minor resistors tweaks (nothing unusual).

Over the last few days, though, I've been combing through some of the comments on some other discussion boards regarding transistor gains.  There seems to be some disagreement about whether the "traditional" values I noted above are what to shoot for, or if the Q2 gain should be a bit higher than Q3.  David Main commented that in one of his favorite versions of the circuit, the gains were quite higher (upper 100s and lower 200s Hfe), and that the Q2 gain was the highest of the bunch.  A few other people commented that they prefer this arrangement too, and that it actually led to less hiss and background noise.

Now of course, I will experiment and see what I like best, but I'm curious as to WHY some people might prefer a higher Q2 Hfe, and why it might lead to less noise.  I am going to guess that part of that may have to do with using a relatively low-leakage Q2 (since it's my understanding that Q1 needs to be leaky to work, and Q3's leakage contributes to tonal character as well).

Curious to know your thoughts.  Also...and for the record...I inherited several Mullard OC42's (NOS).  Used 2 of them for Q1 and Q2 in a Bumblebee (Buzzaround) project with a 2SB176 in the Q3 spot and REALLY liked it.  The OC42's I have  are very, very low leakage, and fall into perfect gain ranges for the "traditional" values for a Tonebender (70-ish, 80-ish, 100-ish), but I wonder if the extremely low leakage might actually not work so well for the functionality of the circuit...