I always thought of its as an increase of inflow from below the rotor disc by exposing more of the disc to the airflow, thus dropping the relative airflow and increasing the angle between the RAF and the rotor tip plane path. As the lift vector is taken at 90 degrees to the RAF it brings it further forward from the axis of rotation thus increasing the autorotative acceleration and the NR.
could post a picture but can't remember how!
Will probably promote discussion, but that's what we're here for eh?
trying this link: