The distance of the piggyback output coil from the rotor and the thickness and permeability of the core are critical factors.

Here are two coils in parallel driving a diametric tube directly with 6 volt A.C. power at 50Hz and 3000 R.P.M.:

https://www.youtube.com/watch?v=9ueoyyVmNk0

Could a capacitor generate a 50 Hz LC tank sine wave wired to the coils strong enough to spin the tube?

Wire a power potentiometer to a six volt battery, then wire a correctly sized "Variable Capaciitor" between the battery and the parallel coils. Reducing capacitance should speed the rotor up with increased frequency. What direction would the power nob turn as the rotor accelerated from increased oscillation frequency, towards a lower or higher input? I forcast we'd gain more speed with a reduction in power!

Impedence matching has been identified as the porch of Overunity.

The wiring schematic is pretty straight forward towards the end of the video.   

Gotoluc shows how a load reduces input while A.C. frequency and R.P.M. remain constant:

https://www.youtube.com/watch?v=TS4zmXU11A8

Here's a picture of four variable Inductors. Two variable Inductors in parallel can power the spinner from the LLC sine wave from a variable capacitor matching impedence. We should be able to control the frequency and R.P.M. by increasing or decreasing the Inductance and Capacitance, thereby altering the resonant frequency.