When I was going through the video I had the above question. Basically, is there a reason we chose Copy Transforms for the DEF-Head bone to the Head bone instead of just parenting the DEF-Head bone to the Head bone as was done in the Wrangler Rig?
I tried out both on the exercise and it seems to produce the same results. The biggest differences that are apparent to me are that the DEF-Head can still move freely unless you lock the axes, but the rotation isolation appears to still function the same.
Because of that my current assumption is that it's just more convenient to do the constraint, (unless you actively need a stretchy rig I guess), though it shouldn't negatively affect any functionality. Is there another reason I missed? Or is there some weird behavior I would need to account for that isn't covered in the exercise?