If we take instead of inductance an ideal lossless electric motor with a flywheel mounted on a shaft. 
that would work also.
but, imagine a ferrite bead used for noise reduction on a computer monitor cable. wind wire on the bead as if you were winding a toroid.
if you bring a magnet close to the bead, it will be attracted to it. but the more current you apply to the winding, the magmet will be less attracted. if enough current is applied to the winding till the core is saturated with flux, the magnet will no longer be attracted to the bead. this was the concept of the second version of the orbo motor by steorn. magnets on the rotor are attracted to the core. when the magnet is pulled in to tdc of the core, we pulse the winding on the core, and the magnet passes by. but that wasnt all. in all that, the winding on the toroid core is not affected by the magnet passing by. no voltage generation, no lenz.
so back to the ferite bead... if we insert a cylinder magnet in the bead that is almost as long as the bead, a lot of that magnets flux will be attracted to the bead as the bead creates a shorter path for the N and S flux. so some, most, or close to all of the magnets flux being in the bead, when we pulse the toroid winding, what happens to that flux in the core? it expands outward. when we disconnect the input to the winding, the magnets flux is again attracted to the core and collapses inward.
so now we wind a second wire around the bead as if it were just a rod core around the circumference. this winding will be very close to 90deg of the toroid winding. if we pulse the toroid winding, there will be virtually no output on the second winding. so, what happens when we pulse the toroid winding with the magnet inserted?

lenzless generator.
so for a single pulse of the toroid winding we get output on the second winding in one polarity when the field expands outward and cuts the second winding and another output from the second winding after the pulse when the magnets field collapses back to the core, cutting the second winding again with output of opposite polarity. 1 pulse input, 2 individual outputs
now, if we have a large enough bead where we can wind a toroid winding with enough inductance to convert the charge from capA 10v to 0v capB, and even just get close to 10v on capB, then we also have the output of the second winding that charges capC using the orbo effect to let the magnets flux expand and cut the second winding to charge capC. and hopefully capC contains enough energy to cover for the small loss inherited by capB plus some, repeat.
back at ou.com we had a thread where we all were coming up with a way to make the orbo motor effect happen solidstate. this was my version.
i wind the second winding first though then the toroid winding. seems the closer the output circumference winding is to the core, the better the output. tried both ways.
mags