I have a friend who does commercials.

He shots DV then he captures 4:4:4 component 10b (from taped DV!!!) and the resulting footage is amazingly good, strong colors, clean.

I tried too and i found that if you don't process the footage, this workflow is useless, but if you do strong cc and processing the D-A-D workflow puts 10b 4:4:4 data into every pixel (including artifacts) and when you cc it, the result is better than cc'ing a native DV codec in the DV space.

Could it be the same with HDV?

that is obvious if you want the best a level can give, you need to work at a level higher.
thus said, if you really want quality, capturing from DV tape to high quality codec is the same that capturing DV to DV codec, then converting in the computer.
and if you really want a good quality, nothing beats direct capture without going to tape compression.