1T308B Russian germanium transistors and R.G Keen's testing method

[j_flanders]

A Russian guy on my local CL had some Russian transistors (P416B and 1T308B) for sale.
They weren't cheap (4 euros for a pair) but he convinced me that I would get some good ones for a Fuzz Face.
To measure them he used that Chinese LCR-T4 ESR SCR meter of which there's a thread on this forum.
https://www.diystompboxes.com/smfforum/index.php?topic=108076.0
https://nl.banggood.com/LCR-T4-12864-LCD-Graphical-Transistor-Tester-Resistance-Capacitance-ESR-SCR-Meter-p-1311439.html?cur_warehouse=CN

They arrived today...
Here's what they measure with RG Keen's method, ( http://www.geofex.com/article_folders/ffselect.htm )
between ()-brackets what he measured,
and between []-brackets what my regular dmm measures (which according to the manual measures at base current 10µA and Vce 3,2V)

1T308B
37 (61) [53]
46 (81) [66]
48 (76) [69]
52 (78) [70]
63 (104)[93]

P416B
33 (68)
41 (79)
43 (79)
48 (87)
60 (97)

Leakage for all of them was in the 20-30µA
But the gains are pathetic.

However, when testing them in circuit they sound really really good!
In fact, I started with the worst pair of the 1T308B (37 for Q1 and 46 for Q2) and even with those I just kept playing for over an hour, not even thinking about trying the others.

So, now I'm confused. I expected nothing but horrible nastiness in every way thinkable.
But even with such low gain it has more fuzz than I need (with guitar volume full up), the cleanup is excellent and the inbetween breakup settings are also really smooth.
This does correspond though with what others have been reporting about these lower gain 308's.

But, if even such low gainers sound so good, why do people hunt down those elusive 70+120 pairs?
And as someone in another thread already said: based on RG Keen's method of testing a lot of very good sounding transistors will be thrown away because they aren't even in the neighborhood of what is generally considered acceptable for a Fuzz Face.

Am I thinking correctly that if you insist on keeping all components in the circuit completely stock (33k, 8k2 bias resistors, 100k feedback, 1k emitter etc) the only good sounding ones are the 70-120 pairs?
That, if you don't mind changing one or more of those, just about anything sounds good/authentic in there?
If so, why not go with these low gainers but also ultra low leakage and stable ones?
I had to lower the 8k2 bias resistor to around 4 or 5k otherwise it sounded choked and a bit sputtery. I guess it'll drain the battery faster but I don't care about that if it sounds this good.

When people decided the only, supposedly, good ones were the 70-120 pairs, how did they test them?

Please enlighten me.

[mozz]

Every tester is going to give you a different reading. Many Russian transistors are tested at higher currents and often will be at the lower end of the spec. That being said, many lower gain transistors will sound great, I don't follow the recommended "gains". But on the other hand, I don't like dual Russian tranny's, maybe in q2 but a Amer/ euro/ Japan/ tranny in q1. I also test for AC beta as some identical "reading" transistors from different makers, are way apart when it comes to AC gains. That's another subject though.