Hello all! I have opened a bug report. Please review and support if you agree.
After reviewing the ground effect behavior for more clues, I discovered that the cyclic is affecting the torque output of the rotor system, which changes lift. With cyclic centered, the torque is as expected. Pushing the cyclic forward increases torque which inappropriately increases lift. Pulling the cyclic back causes torque to also increase. My joystick is only bound to elevator and aileron. There are no cyclic axis to bind to.
What I expect to happen, is for the pitch pulled by the collective (combined with throttle) to set the torque. The torque will remain stable while the cyclic is maneuvered, as it is pitching the rotor disc and not changing the AOA of the rotor blades. Torque is a result of rotor blade AOA & rotational energy from the motor. Lift is a result of AOA & the RPM/Torque Relationship
This means that when you achieve a near perfect hover you loose lift, but with even the slightest application of cyclic you GAIN lift due to the torque spike.
If I am correct, this would explain the very subtle bug of strange hovering behavior & bobbing.
As I am not a 407 SME, I could be wrong on this type. It would be handy if we had a 407 pilot on the forums to verify or dispute this detail. 