I am wondering if the audio issues could have been fixed if the deck could have exported a timecode stream during capture. Have VITC[1] on line 20 of each frame, and the corresponding timecode as an LTC audio stream recorded alongside the audio. I don't know if it help with any drift from the audio capture and video capture clocks not being locked together, but it should reduce the amount of work after the fact to line the audio and video back up, as both streams contain a timestamp of the tape's position.

1. https://en.wikipedia.org/wiki/Vertical_interval_timecode 2. https://en.wikipedia.org/wiki/Linear_timecode