Well, it is not quite that easy. Your leeway sensor idea is a much better idea. Ignoring the obvious problem of current there is an accuracy problem.
The bearing accuracy of a good commercial flux gate compass is plus or minus 2 degrees. The bearing repeatability is typically plus or minus 1 degree. I am interested in measuring leeway with greater accuracy than a 2 degree range, and certainly better than a 4 degree range (really absolute accuracy is what matters comparing to a GPS track). Four degrees is right around the leeway angle I am expecting when the hull just starts to fly on a close reach. I also want to be able to read leeway angle while sailing.
I know comparing course and heading works for finding the set and drift of ships with enough accuracy for navigation. Sailboat performance is a different problem.
