I mentioned the use of pull from a while.
There is only rotors with their respective hollow axes and the rope within, not rod.
I don’t see why we should change the curvature of the tether. Each rotor could have the same angle of attack (AoA). Dealing with different AoA is not needed in this configuration for what I think, even as it can be desirable in another context.
Brainstorming is better than nothing, especially since none of the AWES is close to succeeding on a significant scale. It is a mean to simplify the management, avoiding crosswind complications, and perhaps allowing a better maximization of the space as it scales by the number of small rotors. But some other problems can occur.
Concerning your idea, a detailed explain with a sketch could be useful.
To resume this is like the device below but by replacing a large rotor with numerous small rotors since a rope can be used. One could make the same by replacing rotors with numerous small kites but the implementation of rotors look to be easier as they are settled around the rope.
http://cdn.pes.eu.com/assets/misc_new/peswindissue17talkingpointskylimitpdf-891355291894.pdf
I waited a clear reply from @Ollie to my question about the ratio kite lifter area/rotor area, and he confirmed it on Daisy progress with rigid blades - #75 by Ollie.
So that can be a brake for the system I just described. But I am not still sure of this as the torque is not used as such, only pull.