I have benefited from more than one professor who taught more or less according to the Moore Method. These professors happened to come out of SUNY Binghamton in the early 1980's.
The courses required lots of input from the students. One professor in particular would ask the class to make a conjecture about some construction he would put on the board and then name the conjecture after the student - we were then personally responsible for success or failure of our ideas. After the first few humiliations, this became wonderfully empowering and fun.
The biggest impact I have noticed from taking these courses has been on my independent studying in other classes. I notice that I am able to replicate the types of attacks on my own reasoning that my professor used to - hence leading me closer toward the boundaries of my own understanding.
I don't really know how to respond to your idea of theory building... it seems to me that mathematics is not created upward from luckily chosen axioms, but in a highly nonlinear and back-and-forth fashion. The only example I know of to illustrate this clearly is the development of topology, starting in the late 1800's with Cantor and more or less culminating in the early 1900's with Hausdorff (excellent discussions with references appear in Engelking, General Topology).
I agree that it would be a great benefit to you to find courses which at least approach what you have in mind. Good luck to you.