Glenn Ashmore wrote:

snip
Also by using back EMF to track rudder position limits
the drive train to purely mechanical linkage. I think hydraulics would be
out.

It may work OK with hydraulic drive steering. I've not tried it.

Using back EMF to track rudder position is imprecise for several
reasons, and the errors may accumulate. To accommodate this, there is a
routine in the code which compensates for slow divergence of deduced
position from actual position. It does this by slowly shifting the
deduced helm neutral position towards the running average deduced
position. I think this would also allow slow "slippage" in the hydraulics.