On a gyrokite there are several points in order to change the centre of gravity, as for any kite where the bridle can be modified to change the angle of attack. After all a gyrokite can be only a long stick + a stabilizer + a mast + the rotor whose the weight is equivalent to all other gathered elements. It is the only known way to fly a rotor without a lifter kite or without automated control.
That said a correct angle of attack (until 45 degrees for a glide number of 1 for a large static rotor (not a toy like my rotor), preferably 20 to 30 degrees, and only 15 degrees for a crosswind gyrokite) (see page 45) could allow to increase the total rotor area compared to the lifter kite area.
The maximal elevation angle of my small rotors is about 30 degrees, corresponding to an angle of attack of about 30 degrees when the complete gyrokite is used. But as a lifter kite with an elevation angle of about 40 degrees was used, the angle of attack of the rotor was 50 degrees close to the kite, and still more when the rotor was far to the kite, leading to an elevation angle still much lesser.
So a device to lower the angle of attack is needed, such like I attached the sketch, or preferably the pivot system you describe, allowing an individual adjustment according to the curvature of the line. But that remains in a certain limit (which?) due to the requirement of the lifter kite.
The figure 72 and the figure 73 mention a pivot (26) but I think this is related to a collective mechanism that is not applicable for individual rotors used in yoyo mode.