charlie b said:
But once you have equilibrium established, all that will happen thereafter is that there will be an exchange of energy between the electric field of the capacitor and the magnetic field of the generator windings.
And to add, in order for this to happen, the conductors between the generator and the capacitors will still have to carry this unproductive charging/discharging current, so also must be of zero impedance.
In the real world, we would want the system to be useful, so any current that actually is used at the load end would be in addition to the unproductive current described above; the system must carry both.
When correction means (capacitors for inductive loads) are added to compensate for loads with poor power factors, the unproductive current only has to traverse the conductors between the caps and the load.
If the compensation is exactly matched to the load, the service and supply conductors only have to be sized to carry the real power. Power companies are in the same boat, and add a surcharge to low-PF customers.
A watt-hour meter is basically an electric motor whose speed depends on the strength of both the voltage and the current. It measures volts and amps (VA, apparent power) but it reads out in watts (real power) over time.
If the load is purely resistive, the voltage and current peaks occur simultaneously, there is no phase shift and the PF is 1. If they don't occur at the same time, the motor spins slower, and isn't as accurate.
That's because the system must still be capable of carrying both the productive and unproductive currents, yet can only make use of the latter. Volts is volts, and amps is amps, but they ain't always watts.