This is not what you find in scientific publications, such as those below.
Page 10 and 11 (see also Fig.3):
In analogy to a similar expression in solar power, we might call the loss that is due to the fact that the total aerodynamic force is not perfectly in line with the wind direction the cosine loss.
In fact, all tethered systems need some elevation angle that the tether forms with the horizontal in order to reach some altitude. Fortunately, for moderate angles, the cosine is still close to one, for example the cosine loss is less than 30% even if the tether goes upwards with an angle of 45 degrees. Optimized AWE systems typically fly at even lower elevation angles, and for e.g. 20 degrees we have nearly negligible cosine losses, of only 6%.
The principle is in this excerpt of these sentences: “we might call the loss that is due to the fact that the total aerodynamic force is not perfectly in line with the wind direction the cosine loss.”
See also the equation (45) page 7, which includes the cosine cubed (cos³), below:
This publication is the chapter 12:
Indeed, especially if we consider that two of the authors wrote these two publications including the one I cited last and which relates to crosswind maneuvers.
As you know, to be quick, we take part of the formula to determine the power captured in crosswind flight: Cl (L/D)², Cl being the lift coefficient, L being lift, D being drag.
For a Magnus cylinder, the optimized Cl (with a high spin ratio) is very high, a bit like this. On the other hand the (L/D) ratio is quite low, between approximately 2 and 3, which would give a (L/ D)² between 4 and 9. This is much less than for classic wings. But it is far from negligible. And also the chapter 12 (see Fig. 1.18) and above all the chapter 13, which is available on request, indicate a vertical trajectory far from the winches, and which would be about two times more efficient than the usual oblique trajectory. This vertical trajectory is considered as a crosswind trajectory, although theory (the paper above in the first link of the present comment) indicates that crosswind maneuvers in figure eight are still more efficient than vertical trajectory (about 2 times): this is not expressly indicated but we understand it by cross-checking the information on the two balloons considered in the two respective publications. The problem is that we still have little information (including that from Wind Fisher) on these two crosswind flight modes (vertical trajectory and figure-eight).