With magnetic bearings in vacuum and superconductor for the coil the only energy loses would be the ones that this theory predicts due to mass fluctuations I think?
So either it does nothing and stays where it is, or it converts energy into movement at 100% efficiency?
Electric motors can be ~90% efficient, or even higher for really well engineered ones, so very high efficiencies in converting electricity into movement are common. Electricity is basically movement already -- of electrons -- and is usually generated by spinning conductors and magnets around.
It wouldn't be 100% since some energy would be lost due to internal friction and the need to actively cool the thing to maintain superconductivity (and pump away heat generated by friction). You'd also lose energy in the electronics in generating the oscillating signals, switching, power conversion, etc., and in EMF losses. If it really does work I could imagine it approaching the efficiency of a good motor, just one that creates directional force in a vacuum by pushing against... umm... something. That'd be a question for the theoreticians. :)
So either it does nothing and stays where it is, or it converts energy into movement at 100% efficiency?