Yes, it is correct and can also apply to winches used for reeling systems during power reel-out phases, although, for the lever (wheel) systems, I would say “if the tether is parallel to the axis of the wheel, the power transfer is almost 0, and if the tether is perpendicular to the axis of the wheel, the power transfer of the wheel is maximum.”
Perpendicular to the axis of the wheel or “parallel to the take up wheel” (for lever systems the two expression can design the same as mentioned above) leads to an optimal transmission without restriction of length. Note that the same applies to the transmission of a reeling yo-yo kite during power reel-out phase, and with less materiel and greater adaptability: regardless of the elevation angle, the tether is always “parallel to the take up wheel” (to take your words here quite suitable unlike my expression (“the tether is perpendicular to the axis of the wheel”)) at the point of its exit from the winch.
For TRPT like Tiira, things are different: the tether angle is between 90 degrees and 0 degrees. So the transmission cannot be optimal if the tether length is without restriction. But TRPT work also by the axial force (tension by the power kites or blades, and the lifter kite if applicable). Under tether length restriction and a high axial force, the torque transfer can be maximum. If one imagines an unlimited axial tension, then it would function like a rigid structure for which perpendicularity would have no importance.
Now the calculation for TRPT is complex because of a lot of parameters: the axial force, the torque, the length of the tethers, the number of tethers, the phase lag (expression that I had forgotten and that Tallak reminds me of)… It is not “tether angle losses” as such, but among parameters that constitute the TRPT, knowing that it is about determining the maximum possible distance between the wheel and the rotor (blades or kites) while having maximum efficiency or almost.
Said like that, I can only agree, and that’s what I suggested, just like you from the very beginning.
In the end we can agree that the rope-drive transmission allows avoiding tether length restrictions while maintaining optimal efficiency.
More and more, I will lean towards a traditional wind turbine or rather a Kiwee-type because lighter, at least concerning the rotary devices, separating (just like you suggested if I remember correctly) the lift function from the power generation function.