Herein lies the problem. Your proposed solutions are ever more complicated as you begin to grasp the magnitude of the issue. Implementing this idea, and I'm not saying that it's good enough, would require modifications to pretty much every major telescope in the world. And there's no economy of scale because most of them are custom builds. Who's gonna pay for that?