The waterwheel analogy is apt.

You should, in fact, be able to build a very similar yo-yo device to replace a waterwheel, though it would likely be only drag powered, with no lift-generating mechanism. Makes me wonder if there is a yo-yo application for tidal generation, with the direction of the kite shifting with the currents.

It is more than a waterwheel. Consider a high performance sailplane wing with a 60:1 lift / drag ratio. It could generate vastly more tension in the cable than a simple kite of equal area.

Back to basics. Air has mass. Moving air has kinetic energy. Wind machines extract the kinetic energy by slowing the air down.

Consider a modern windmill with thin long airfoils at a standstill. The blades intercept perhaps 10% of the air flowing through its disk resulting in a slight reduction in average speed and a modest drag force on the tower. Now spin it up to maximum power conditions. The airfoils slow down all the air passing through its disk producing a much larger load on the tower.

It is that ability to absorb energy from a large mass of air with a small device that is the difference. If the concept succeeds the kites may evolve into rigid wings because their high velocity allows them to absorb energy from a much larger mass of air.

I think the show stoppers are;

1. Large land requirements per MW.

2. Large airspace requirements per MW.

3. Noise. A high performance sailplane at minimum sink speed 3000 feet above the ground is silent. But at high speed it is audible on an otherwise quiet day.

Consider a 1000 lb sailplane at high speed sinking at 250 ft/min. it is converting 7.6 hp of potential energy to maintain that speed.

A 3MW rigid wing machine would be equivalent to 531 such sailplanes flying in formation. They would make considerable noise. A 3MW kite would be slower, much larger and would produce much more noise.

With increasing wind speed the required area would shrink as the wing loading increases, but I would expect the noise level to stay about the same.