For practical purposes you need to consider energy returned on energy invested not just power flow rates, although power flow rates can be an issue. Less cost and less environmental damage to just use a battery.

A simple optical sensor could sense rotation, and there's probably some weird meteorological situation or chemical plant process meter that would find a rotating optically detected non powered humidity sensor to be useful.

For chemical process plant work something that measures humidity would be super boring because there's a zillion competing technologies, all mostly unpatented by now. The fun would begin with selective uptake of "weird stuff" non-water substances using the same physical design. Maybe you could detect H2S leaks or hydrocarbon leaks or ...