Brush topology is not good because of possible different winds in the area. Adding network aloft would make the things still worse, by adding constraints linked to supplementary ropes increasing the risks of tangling.
Photos of networked kite are nice, but these kites produce only force, not power. Also some architecture called as network can be in reality a single flying member comprising several elements, not a network as such. For example Beaujean’s AWES could be seen as a network as it contains numerous elements but it is a single flying member. One could make the same remark for an arch like Mothra. Indeed some version of Mothra could deserve to be studied in a powerful (yoyo?) version on a ring allowing facing wind direction, perhaps like on
See also Completing Mothra, including variants.
It has happened before, so it will happen again. The same for computer failures.
Wind changes can occur in a farm area. An example is when the main wind is changing: a kite can be affected before the other kites, leading to increased risks of collisions.
“Perhaps” is not enough. AWES are expected to work 24 h/24 all the year, meeting all possible conditions.