I may be wrong, but isn't the only type of rotor where the disc actually "tilts" a gimbal rotor? With other rotors, the cyclic only varies the direction of the rotor's net thrust vector, right? While the collective varies the total thrust (or lift) from all the blades along the vector controlled by the cyclic.
I also don't believe that the cone angle of a rotor's blades technically changes. The blades may experience more bending deflection from root to tip as the amount of lift each one produces increases, but the cone angle at the blade's hub attachment will remain essentially unchanged. As cyclic is applied, the lift produced by any given blade varies during the course of a single rotor rotation. And it's the asymmetric lift condition resulting from this that produces the cyclic directional control. The cyclic inputs also result in each blade experiencing a varying amount of bending (flapping) as it goes thru a single rotation.
As for the situation where the aircraft must auto-rotate, I think we must consider the difference in a rotor that is being driven by an engine, and a rotor that is relying on the free-falling/forward motion of the aircraft to produce sufficient airflow over the rotor blades to spin them fast enough to create enough lift to slow the aircraft's descent rate to allow a safe landing. Basically, the direction of airflow thru the rotor during powered flight is somewhat opposite of that during auto-rotation.