Thanks Lord Tim! I look forward to trying that. Interesting: a clean gated track (s) is necessary to generate a model for the other tracks to "hear" after which quality is upgraded by an additional move. Is that the gist?

Tried to reply last night but the forum kept eating it. Sigh.

Yeah, you're really only using the gate tracks for detection, they have no bearing on the sound quality at all.

The thing that does is when you quantize your tracks (is: the ones you'll actually be keeping) you'll need to bounce them to clips so it both freezes the transients and uses the superior offline stretching algorithm.

Ok Tim (thanks again for the details!) Yea... I understand your first point re: gated just for detection.
Question: In your step 4, am I bouncing both the gated snare and gated kick to a single track or to 2 separate tracks?
It looks as though step 9 accounts for phasing issues between e.g., snare and overhead?I may have additional questions this weekend when I try it.

Yeah, you need the markers to be at the same place across all tracks and move exactly the same, which is the reason it's all added to the pool and the pool is quantized, or yeah - you get terrible phase issues everywhere.

You really don't need to create separate detection tracks if your drum part is fairly simple - most rock stuff is, most prog metal is NOT - you should be able to get clear hits on kick and snare without them getting overlapped or confused. The other reason you may want separate tracks is to do transient to MIDI conversion so you can overlay drum synth samples. If you don't need to do that, save yourself a step and bounce both your gated tracks down to one.