Lu,
You said "When in a zero g condition the rotor is unloaded and therefore, difficult if not impossible to control."
As I understand it, this is not true. The reason zero/low-g is dangerous is that while flying at positive g, the helicopter 'hangs' from the rotor like a pendulum (yeah, it's more complicated, but the point can still be made). When you push over and get into a zero-g condition, the fuselage is no longer hanging--the rotor no longer imparts any significant forces on the fuselage. Now the major forces acting on the fuselage are the tail rotor and any aerodynamic forces due to positive velocity. At this moment removing the main rotor would have no effect (I take that back--there's a lot of torque, but that only plays into how much T.R. force is necessary). The fuselage rolls right because the tail rotor at this moment is above the c.g.
During normal positive-g flight, if something causes the helicopter to roll right the pilot instinctively nudges in a little left cyclic. Since the rotor is still being controlled by the swashplate (and hence the cyclic), doing so will cause the rotor disc to tilt as would be expected. But because the helicopter is not 'hanging' from the rotor, it will not follow. It will continue rolling right. If the angle between the two becomes too great the area around the teetering hinge will come in contact with the rotor shaft (or a rubber bumper if it's attached), and/or the blades will contact the left side of the fuselage.
I don't quite understand how slight right cyclic is bad. It seems that a combination of slight aft cyclic to reload the rotor and make the helicopter 'hang', and slight right cyclic to ensure that as the helicopter fuselage rolls right the disc follows (thus ensuring that the disc plane remains near perpendicular to the rotor shaft), would be the best way to get out of a low/zero-g condition. Even if you end up doing a positive-g diving right turn with very large (>>60°) bank, you're still ok. How is this unrecoverable?