The document below allows a more accurate evaluation of the potential of a crosswind gyroplane, above all the figure page 45 I mentioned on Bryan Roberts' Gyromill (Sky WindPower) topic:
The values concern a complete and not recent gyroplane. Certainly the lift-to-drag ratio / rotor plane angle of attack would be better with a recent rotor and a thin fuselage that does not require the carriage of passengers.
In the meantime we are going to do with the graph data. The angle of attack (AoA) of 15 degrees leads to a lift coefficient of 0.6, and a drag coefficient of 0.2, so a glide number of 3. As an approximation and as a yoyo system our 1.7 m² rotor of 8.4 m diameter sweeping 55.4 m², wind speed 10 m/s, air density 1.2, 30 degrees (power = 0.65 after cosine cubed loss) of elevation angle would lead to: 1.2 X 2/27 X 0.6 (1000) (9) X 55.4 X 0.65 = 17285 W.
This relatively high number, even at a low glide number, is due to the large area of the rotor disk plane (55.4 m² in the example) compared to the rotor area (1.7 m²), knowing that the lift coefficient of 0.6 is a not too bad value in this context. An advantage can be the low travel speed compared to the pull, maybe allowing more safety, a possibly easier management of multiple unities as dancing kites (Y configuration), and less tether drag.
If the system is a flygen, the apparent wind would be crossed by a rotor at 15 degrees of AoA, leading to huge losses (squared cosine?) compared to Makani or FlygenKite turbines facing the apparent wind. So a flygen crosswind gyroplane does not seem to be a viable solution.