Adam, there was a good thread on this which i can't find right now. S76, among others, does vary Nr with speed, load factor, and height. It is becoming the norm now.
The headache is that the rotor blade eigenmode frequencies are a function of Nr. Since many orders of P may be aerodynamically excited, and could work there way back into the structure and systems, the machine must be carefully evaluated at different Nr. Worst case is that blades vibrate in phase to compromise fatigue life. The other problem is that the azimuth lead angle is a function of Nr, so you need additional complexity in the control system.
The golden panacea is the ability to vary rotor Nr from 100% all the way down to near 0%, in an advancing blade concept helicopter. This means that while efficient hover can be achieved, the machine can have a cruise factor unlimited by advancing blade compressibility. The limitation is the need for a heavier rotor system to transfer the root bending moments.
I am a very keen fan of the Sikorsky X2 project, although do not know it's current status (scrounging for resource probably).
Mart