@buherator Yesss. I've idly pondered about this a while back, some kind of goal oriented system, where certain visibility and alignment constraints contribute to some kind of utility score.
Not sure how tractable it would be in practice tho.
Maybe a greedy approximation would be enough?