In your experience, what range of azimuth changes resulted in optimization. 1/8 turn? 1/4 turn? Or less?Much less. 1/12 or 1/24 of a turn. It is important to take up the backlash in the threads of the adusting screw BEFORE loosening the two set screws. Otherwise you'll have no idea where you started or how far you've actually rotated the armwand.
Also, did your setting through listening equal the setting you obtained using the Wally instrument?Yes, pretty much. With the UNIverse I was able to get crosstalk below 0.5db. Subsequent settings by ear seemed to duplicate that quite well. As other components in the path get better, it becomes easier to hear and adjust accurately. Nick's components and the Valkyra wire really made it clear.
Finally, did you have to connect your speaker cables to the Wally device or can you just insert tonearm out puts or preamp outputs?Speaker wire connections work much better. Low level signals like preamp or, especially, tonearm output would require much more sensitive measuring equipment. The Wally just uses a standard digital multimeter. No way is that sensitive enough to measure crosstalk accurately with low level signals. Channel imbalances in the amplification chain don't matter, so there's no reason not to use the more easily measured output from your amp(s).
Doug
P.S. I don't bother levelling the headshell. That would be useful if the cartridge were dead nuts on from top surface to stylus, but that's unlikely. I just start by making the stylus/cantilever look vertical (under normal VTF on a record). This seems like a logical and simple place to start, at least to me. I noticed Frank Schroeder doing it the same way. Diff'rent strokes?