The real issue with an adaptive surface propulsion control system is what you have to know how far in advance to allow the system to respond properly. Using a ground effect vehicle as an example, it is feasible to make an attitude controls system that senses the ground and obstacles ahead and either determines that they are overflyable or avoids them. With modern ranged ground sensing LIDAR merged with sub-millimeter radar, LRGA, and a MPP backbone server, doing this on the ground is possible...think advanced DARPA ALV. But as I said, ground vehicles are dirt simple, and that is because it is dirt...static...it is not changing in the few seconds before you get to it. Waterborne craft, especially those that have to interface with the wave profile, have it harder because the surface is changing in elevation and slope constantly.
Because of this hulls are designed to interface with the surface in one of 3 ways, profile, platform, or mixed response. Mixed response is what a normal buoyancy supported hull does; depending on the wave encounter frequency the hull either responds to the wave profile (waterplane dominates: a barge) or doesn't (mass dominates: SWATH). Platforming and Profiling are what high speed and most dynamically controlled vessel do. In Platforming, hull is controlled at a fixed height above the still water surface with very little vertical motion either by design in planning hulls, air pressure in hovercraft/SES or by control surfaces in hydrofoils. Once the waves reach a certain height and length, it is possible for high speed craft to Profile where the hull is controlled to follow the swell surface with considerable vertical motion.
However, craft that rely on Platforming or Profiling usually cannot transition from one to the other. There will always be a point where the wave profile cannot be mitigated by the design or the control system. It becomes a matter of being able to sense the wave surface with enough fidelity in enough time to respond. In the design postulated, even with the best modern sensors I know of, there would always be some randomness that could not be sensed far enough in advance to move the control/propulsion system to the optimum. This problem is very close to what some of the AC teams were trying last year with the wind.
http://www.sail-world.com/Cruising/...-Oracle-Racings-wind-mapping-technology/66144
Some day there might be the capability to get a good enough estimate of exactly where the water surface will be at the propeller the necessary time in advance. I know that in my own experience with real time sensing systems we had to wait for computing power to catch up with the theoretical concept.