Whereas I would not dream of doubting the beautiful mathematics of the technique, the whole thing falls apart when you question what 'flare' is.
You flare at xx feet. OK, but at what rate do you flare and how much control force is required? Throw in a few more variables such as weight and C of G position and you're back to square one.
Obviously, what Mr Jacobsen has done is to plagerize the autoland logic and programming and substitute the autopilot with a human pilot.
The software for the autoland is such that the desired result is programmed in and the autopilot adds whatever control inputs required to produce this. The computer knows, (or at least, can calculate so quickly that you wouldn't know that it didn't), what force and rate of change of elevator deflection will be required to 'flare'. We can't, and must rely on visual and tactile clues.