An Interesting problem which is likely to increase electricity prices.
On the other hand if demand is low wouldn't we expect the electricity market's 'spot price' to be low too? - so even with a small payment to one or two generators to cease, the overall costs aren't necessarily increased. The nice thing about wind is that it is reasonably 'switch-off able' - unlike conventional thermal stations.
I do not really see why it should increase prices - it may mean that there is more to be learnt about managing a grid with non-synchronous generation, but those lessons need to be understood anyway, as we make more use of electronics with inverters at the generation end of things and VSDs at the load end, there will be precious little that cares about the exact frequency.
As I'm sure I have said before it is quite possible to make an inverter with load sensitive timing, that emulates the phase/ frequency shifts that an inertial system has, it is just that right now, generally we do not, and a lot of DC - AC stations on DC links produce very non-sinusoidal waveforms and rely on being diluted by other sources.
Take some of it with a pinch of salt the 'wattsupwiththat' website needs to make a good article out of it. I'm not aware of wild frequency fluctuations, large harmonic currents, or rolling blackouts, more that the existing admin process makes certain assumptions about what generation can be wired in parallel without the system becoming too bouncy, and these assumotions are being tested.
