It would be good to add an option to choose which wind speed to use (default or air density corrected) in the power curve analyser widget.

You can see the power-curve-model based on the anemometer and power measurement as a first indicator.
I think we all know that the p-c-model is not the real deal. To measure the real p-c-model you have to put the whole turbine into a testing-bay (that would be a little bit expensiv). There is also the possibltiy measure it based on the IEC 61400-12-1 with met. mast and build the transfer function
or based on the IEC 61400-12-2 with the anemometer (to build the transfer function you need also an reference near by the turbine... like a met. mast or another meteorological station). Only the first option would matter the OEM (IEC 6140-12-1) when you try to claim.

Anyway, to have an "first indicator"-p-c-model is great and improve the model won't hurt.
Just by the way:
There should also be a transfer function running by now on turbine-pc. The transfer function is intended to reduce the effects on the measurement, caused by the turbulence of the rotor and so on. Makes the measurement better but not so good as from a met. mast.

This is idea is mostly pointless as wind speed data are measured with anemoters placed behind the rotor and therefore heavily disturbed. Trying to improve the power curve accuracy with air density will lead nowhere because the wind speed is too much unprecised.
The only way to have an unchallengeable power curve is to do an IEC power curve measurement with met mast.