Originally Posted by
rpenmanparker
I would think sensors on all sides with the appropriate software to do the additions and subtractions that would yield the correct real and apparent wind speeds and directions.
You might make it work with only two, but you also need to know the ambient air pressure. Many air speed sensors are designed as a kind of fluidic comparator for this reason, with one pressure plate in a still area.
I've kicked this around with a buddy in the context of avionics and yes it's possible up to a point but it gets tricky real fast for any kind of precision. I understand one of the attractions: with speed, weight and grade you could get a power estimation (since there are experimental methods to calibrate for drag-area). There is a relatively cheap product which purports to do this. But the devil is in the details, and it's not so easy to get enough precision that the results aren't thrown off too much to be useful.