In feet, one would hope
Actually feet per minute.
although wouldn't use it above about 50 degrees as it becomes very inaccurate, but up to that, very close to the sine values!
Above 50 degrees you use the full wind strength for the x-wind component. That's actually very close to the sine values as well. In fact, you'll find that the inaccuracies of this method are less than the inaccuracies caused by the wind being variable, particularly in strong winds where it matters most.