I've been playing with the fuzz face bias resistors and found out that under certain conditions both base currents can be identical. At this point the fuzz is a litlle, little better... well, at least for my taste.
Refer to http://www.geocities.com/guitarfxs for a mathematical analysis of the bias resistors.
If someone have an oscilloscope, and the time, to check this I would welcome any data.

By the way, always use a trim pot instead of the 33K resistor. Sometimes the best bias are near 20K or less, very close to the point where the sound begins to fade out.