OK forget the gyro ~ for a moment.
Another idea added to the stack.
The following regards a 2-blade coaxial configuration that incorporates any means and amount of 'rotor rigidity' (IE. not a simple teetering rotor).
For a given amount of cyclic input, it appears that the amplitude of the vibration will differ when the direction of the cyclic input is the same as one of the blade crossing azimuths from when the direction of the cyclic input is midway between two of the blade crossing azimuths.
It therefore appears that the vibration may be lessened if the blade-crossing azimuths can be actively varied so that they always have a constant relationship with the ever-changing cyclical input azimuth.
This coaxial transmission should be able to change the crossing azimuths.
Just a wild idea.
Dave