I think all it would need to do is apply downward acceleration, mass of the board, mass of the user on the board, and use kinematic equations to figure out how much force it needs to exert in the opposite direction in order to remain exactly where it is elevation wise. Like you say though it would also need to figure out the distance above the ground somehow. The mass of the user can be determined by putting small springs in the board itself and measure before takeoff. There are other variables since it's not operating in a vacuum but I'm sure they can be identified and accounted for as well with more in-depth analysis.
Yeah, I took some time to think about it. Wouldn't even need to go that far. Just make the controller a diff-eq that outputs energy using distance, velocity, and acceleration and you'd have the correct control scheme. Give it a variable forcing forcing function and viola.
Awesome. Now all we have to do is work out this anti-gravity crap and we really can put out a product like this. And by we I mean you, Ill stand here and supervise. We'll be rich!