I think it is perfectly reasonable to drop some CPU architectures that haven’t been relevant in the last 20 years. It seems to me there are a lot of new people eager to contribute who have no interest in touching C any more than necessary, and a project that can no longer attract new contributors will sooner or later die.