Whatever the benefits or otherwise of being rough or smooth on the controls, hand flying without too much help keeps your scan quick, plus you quickly know what to do when performance isn't as expected.
By all means embrace all help given by automation, but even with extremely reliable equipment, crashes still happen because the pilot(s) monitoring don't take positive, appropriate action.
Why not? I suggest because they get too used to expecting the gear to work, as it almost always does, and are naturally surprised, confused and slow to react when things go wrong.
It's fixed wing, but an airliner crash on an investigation show I watched the other night is a great but tragic example - 3 pilots on the flight deck, coming down the ILS, captain's U/S radalt tells the autothrottle they're about to land (although they're nowhere near it), autothrottle pulls the power and they stall in from a few hundred feet.
The question I immediately asked myself was, where was their basic instrument scan? Why didn't at least one of them wake up to the looming power / attitude / performance disaster and jump in with a manual go-around?
Hindsight is great and I've stuffed up as many times as most people, but that fast connection between eyes, brain and manual control movements engendered by hand flying should always be maintained - IMHO!
However it is not a perfect carry-forward. Surely it would be much better to have the training optimised to the current skill set (including of course monitoring and situational awareness), rather than the outdated one and just hope the carry-forward is adequate. Training time is a limited resource, it should be optimised to the current role and not one from 30 years ago.
Yes, fully agree here - we must also move with the times, not get caught saying it was better in my day all the time.