I tested three torus (see in the preprint): a thin 36 cm tire, a 42 cm tire (internal diameter 24 cm), and a 85 cm buoy.
The 42 cm diameter tire was better due to its proportions, neither too thin nor too fat. The rpm of the propeller was 1.1 to 1.3 higher in the tire (measurements by rpm-check after those by anemometers). If the cube of wind velocity applies as expected, the power would be between 1.3 and 2.2 times higher, which is not too bad for a simple thing which is primarily a protection for the turbine, and which would allow easier stacking of the turbines.
Note: where the torus (tire) was tilted, even little, its efficiency dropped drastically. Thus, I don’t see how a longer tilted tube could be efficient in regard to the wind turbine inside.
They apply formulas, only seeing the problems afterwards, realizing then that the formulas were not complete.
You will soon see that what is called a crosswind kite (including FlygenKite) is not so advantageous. If you add the weight, that becomes impossible.