Pprunefan,
Actually, we use the term dihedral to describe the effect (when a yaw makes a roll) and can be measured on airplanes, helicopters and even ships! the paper describes wings but the effect is the same.
For helicopters, lost of stuff happens in the rotor, as you describe, like flaping and stuff. The rotor dsk is also coned just like an airplane's wing, so there is a contribution there as well.
But, helicopters generally have strong dihedral due to the way the rotor behaves when we impose a new wind direction on it. If you can, picture the rotor having been trimmed with the cyclic forward at high speed, so that the disk is level and the nose is down to trim at that speed. Then simply rotate that helicopter a bit to the right, like a right yaw would do. If you look at the rotor from the wind's perspective, where the air meets the disk as the new forward, you can see that the cyclic is now trimmed to the right relative to the airflow, so it has a small roll input. this makes the aircraft roll in the direction of yaw, so-called positive dihedral.
the problem with helicopters is that they fly at very low speeds. By the time you slow to 50 or 40 knots, the small effect I describe above is lost in the mud, and there is little dihedral for small angles of yaw. This is often the problem that sets the low speed Vmimi limit for helicopters. that big fin on the rook probably helps get some traction at low speed to add some roll, and perhaps gets the dihedral effect back at 40 or 50 knots, allowing Vmimi to be set lower. VFR aircraft don't need it, of course, since that dihedral requirement is an instrument thing.