Talk about completely missing the point. You were off to such a good start when you said, "Presently, this feature of X is being under-utilized." But from there, you should make the jump that nobody gives a damn about a network-transparent graphics protocol and that every minute spent thinking about how to make it better is a wasted minute. The solution is not for application writers to better support X's model, but to realize that X's model is without value and scrap it. Why are you so interested in improving a feature that nobody uses or wants?