Very interesting new observation on the lip-sync problem when using external HDMI input...
I had my Panny 65VT50 calibrated today by Jeff Meier, who apparently has quite a bit of experience working with Oppo players. I brought up the subject of the 103 sharpness 0 vs. +1 for discussion with him, and although he wasn't familiar with the contents of this thread he did have his own intuitive feelings on the subject, i.e. that 0 should of course be the optimal setting. He was quite interested to see visually what things looked like on the VT50 (sharpness set to 0 there) both ways.
So he then spent a fair amount of time looking at several test patterns, going between 0 and +1. When it was all over hIs final conclusion was that 0 is what he felt was correct, and +1 is not. He didn't feel that the Oppo "star field" sparkles (under the logo) was a valid test pattern, although he granted that it did become "sharper" at +1.
Anyway, while he was working on the 103 for this issue he reviewed the rest of the settings I'd installed to eliminate anything else the player might have been doing or adding, that was unnecessary and/or undesirable (in his mind). In particular, he changed Video Setup -> HDMI Options -> color space (HDMI1 and HDMI2) to YCbCr 4:2:2. This was different from the YCbCr 4:4:4 which is the recommended setting expressed at the front of this thread, and is also the value returned if AUTO is set (since the 65VT50 supports that).
I asked him why he was doing that, and he said that "there is no 4:4:4 source, so the player is simply padding the 4:2:2 input with zero bits to produce 4:4:4 output". By setting the output to match the input it was that much less that the player was doing to process the input in some way.
Ok. He's the expert, and if he says that's "better" (or at the very least not harmful) I'll buy it. So that's how I've been running ever since he finished the calibration early this afternoon... YCbCr 4:2:2.
Now I bring this up because ever since this change was made, I have NOT SEEN THE LIP-SYNC SYMPTOM from any of the HDTV programs I've watched today (from the Linksys DMA2100 WMC extender running through external HDMI input to the 103). This is unheard of in my experience.
I will obviously now pay close attention going forward, to see if this "giant breakthrough" continues, or if I was just lucky with today's viewing.
I'm still using downmix-to-stereo analog audio going from 103 FL/FR out to my analog L/R input on my AVR->speakers (because I thought avoiding HDMI audio might fix the lip-sync problem, which it did not), and things seem to be working perfectly. Lip sync problem seems to have disappeared, and the only change is going to YCbCr 4:2:2 so I have to attribute the fix to that settings change.
If the lip-sync problem disappearance holds up I will try going back to HDMI audio from 103->AVR, just to see if the fix is independent of the type of audio output and seems tied to the YCbCr 4:4:4 setting, fixed by going to 4:2:2.
Just one more data point on this subject. Wanted to share this discovery with others.