I just built then experimented a rotating reel system comprising a ground ring with ropes and pulleys, rotating a superior ring with my hands. The conclusion is clear: when the two rings are aligned and parallel (not rotating reel) the rotation is far easier than when the superior ring is tilted, involving ropes sliding in the pulleys forth and back (rotating reel). It becomes yet more difficult when the superior ring is both tilted and distant, as expected with the figure 22.18. In the other hand my pulley system is not multi directional, so their cyclical adaptation to the angle of their respective tethers is not assured. So significant frictions occur. However I will provide a possible explain in the dedicated topic. However other (and this time simpler) methods avoiding a compression structure can perhaps be possible.
However the chapter 22 mentions in the page 566: “This essentially means that the transmission efficiency for the ideal system in steady-state operation is, as expected, 100%.” and in the ages 564 and 565: “The solution is approximative because the residual transverse forces are causing compensating motions which are not taken into account in the analysis. However, the following results indicate that the effect of the compensating motions is minor and can be neglected.”
A 100% transmission efficiency is theoretically possible. But the physical model is not respected by using leashes (different forces during winding and unwinding) or by using pulleys (forces theoretically equal and constant if the pulleys are well oriented during the operation, but it is not the case). There are also frictions.