LW and IFMU-your thinking brought back some history.
My first thoughts went back to the rear seat controls in the CH-54 where the cyclic controlled pitch/roll and yaw. But the results of that effort were certainly muddied by the fact that the inputs were fed into the AFCS and thus had +/- 10 percent authority limits and the back seat pilot had a cross pointer indicator telling him where he was with regard to limits ( and yes-I recall the B model had the further
RH-53D AFCS where the trim system extended those limits. The challenge was thaat when one had to use a good amount of that authority, associated with good angular displacement of the rear cyclic, it became hard not to unintentionally make a pitch or roll inadvertent input when over at some substantial twist angle of the cyclic due to a crosswind, ot any combinations of the same.
Some of us protested against doing that for the Comanche but I have to admit they minimized that sort of problem. ( Comanche did retain a separate collective control )
There was, at the time some efforts ( was it the Army NASA community? ) discussing the use of a single control inceptor, but in any case when it came to a decision, the FBW S-92 MHP Canadian ship came out with electric pedals and collective. Project pilot for that effort was Rus Stiles, who also had been the Comanche project pilot. He is a gifted aviator and I’d invite Rus to comment on the challenges of a single control. it would also aid this discussion if some of the commercial S-64 pilots who perform such miraculous work could add their experienced thoughts.
In my own thoughts on the subject, issues associated with max slope landings, auto landings, gunship helicopter operations, precision load placement operations, height-velocity/Cat A operations testing would seem to present challenges to precise flight performance due to inadvertent control inputs in one or more unintended axis.