Probably because moving from one high population server with a queue problem to another high population server and now givign them a queue issue isn’t the intended goal. They are trying to distribute that higher population among the smaller ones in hopes of just not moving the population issue from one cluster to another, only for people to turn around and complain about it again.