blcl - when I learnt to fly on a "real" helicopter (Bell 47) I recall that during early attempts at flying, something like 95% of my attention was on controlling the engine rpm with the twistgrip throttle so as to keep it between 2950 (bottom red line) and 3050 (top red line) - though really it had to be kept between 3000 and 3050. This was a significant part of learning to fly the helicopter, but now that skill is completely redundant since I fly a heli that doesn't have a throttle of any description.
However there are plenty of new skills I had to learn relating to it such as optimal use of the automation etc.
It's true that one has to maintain basic skills to some extent, but as the automation becomes more reliable / has greater redundancy it becomes less important.
I think fixed wing world is slightly different because you do fly your stable planks manually from time to time. With our unstable helis, you never fly without the autopilot with pax (they would be sick!) and just have to demonstrate that you can fly without the quadruplexed autopilot in cruise / transition to hover and landing since even with the redundancy, you are only one wrong button press from dumping the autopilot.
However, there has never been a requirement to fly an ILS without the autopilot engaged and as far as I know no operator of large IFR helis requires this (the CAA certainly don't). The requirement is to fly an ILS "manually" ie with the basic autopilot engaged but without the upper mode coupling. In older aircraft, there was a separate coupler box so it was quite feasible to be able to have the basic autopilot functioning with the coupler box not working. In the 225 there is no separate hardware that is used when the approach is coupled vs manually flown with basic autopilot, so the condition where the autopilot is working normally but you can't couple is just not a realistic one, so why take up valuable training time in practising it?
HC