I would add that starting up windmills (getting them started) costs energy and one may want to optimize the decisions taken to start them up. E.g. if the wind forecast predicts good wind only for half an hour or so, it is probably not interesting to start it up. This is especially true for vertical windmills, which are more efficients but require more start-up work.
Also, in the long term, the forecasts may allow better location of windmill farms and their positioning.
So, it's all about the money. Listed so far:
- optimization of start-up, possibly allowing the construction of more efficient plants.
- propagation of forecast of power delivery
- prevent damage
- improved location and positioning
Just my 2 cents,