Roger,
I'm having trouble following what you're saying - and I
really want to understand it!
I agree that it's much easier to see tonal relationships in B&W. And I follow you as far as having two copies of the image - one in RGB and one in Lab.
When you say that you
"look at the different channels and adjust the channels with layers or curves or blend them with channel blender without regard for the color shift. Net result is to combine/blend/mask the b&w channels to make the best b&w image." I'm not quite sure I understand what you're doing. Are you working on the channels in
both copies? I.e., the RGB and Lightness channels? Or only on the L channel? Though when you say "without regard for the color shift", it sounds as though you're working on the individual RGB channels.
OK, now as I'm reading this for the umpteenth time, it sounds like you're working with all possible information to make the best B&W possible. You choose to do this with various blending modes and masks on a combination of the RGB & L channels - but essentially any method that gives you a B&W that you're happy with will work, right?
Then, use that B&W to layer on top of the original image in luminosity mode to in effect apply the tonal separation you just achieved in the B&W.
I completely agree with your observation in #2. I have had to work to restore an image first in B&W, then add the color after the repair was done more times than I would have liked. But, I have found that my results are better with that technique than any other I've tried.
OK, I'm going to try to find a photo that will fit situation #1 and see if I understand as much as I think I do.
Thanks for this tip BTW!
Jeanie