In the video you can see that the ring with balancing balls is situated at about the same radius as the imbalance mass. If you would balance a windmill rotor this way, the ring with balancing balls would be situated at about half the blade length. This seems practically impossible to me. Another problem is that the balls all move to one side if the rotor stands still or is moving only slowly. So the imbalance is not balanced at low rotational speeds and this imbalance requires a rather high starting torque of the rotor. I think that this way of dynamic balancing is only possible for fast rotating disk shaped objects but for those objects it is a clever way.