Although it’s normally bad form to quote private correspondence on a forum, I think it’s fair to quote helpful correspondence from UDO tech support here, perhaps lightening their future load. Here’s how Mike at UDO explained it to me, which is clearer than any paraphrase I could offer:
“Yes, I was talking about the LR PHASE control in the LFO 1 section. When in non-binaural mode, this fader allows you to control the amount of pan spread. At its highest setting ( 2π ), it will cause the voices to be hard-panned to the left and right channels, which is the optimal setting for avoiding any phase cancellation when you’re only using the left output for a monaural signal.”
and here’s his comment about the LFOs and PWM:
" I should have been a bit more precise on PWM in non-binaural mode: Although you essentially have 6 LFO 1s for all 12 non-binaural voices, you will still be able to modulate DDS 2’s pulse width by using LFO 1 as a modulation source. The only case in which DDS 2’s pulse width will be modulated every other voice is (a) when you use LFO 1 in ONCE mode (one might say that this is the option where it becomes obvious that there are only 6 shared LFO 1s for 12 non-binaural voices) or (b) ENV 1 as a modulation source, since the mixer makes use of 6 pairs of VCAs. If you combine the latter with the fact that DDS 2’s pulse width does only have one control per voice pair, you end up with PWM on every other voice only."