I guess the way I can explain why I don't want theoretical applications in my theory class is "That is what you are going to be doing in every other class." In other words, every paper for every other graduate class will take some text as its basis, and also have some theoretical framework. What I want my students to do is isolate the theory for a while, work with it by itself. If I am grading a theoretical application, and there is not much theory in it, then it is hard to give it a grade. It might be a perfectly good paper, but one in which the theory does not work at all, for example.
So the theory course is the one place where theory cannot be "used." Does that make sense? Because one common way of using theory is to name-drop, but do an analysis that could have been done the same way without that theory. To his credit, Paul Julian Smith doesn't use theory in this name-dropping, redundant way. He really does apply the theory in a thick layer.
Also, I like my theory kind of subtle in an actual paper of literary criticism. I cannot say "apply the theory, but make it subtle," and then turn around and say, "where's the theory?" If I ask for an application it cannot be a subtle one.