I think this is the part that often gets underestimated: school reform has to work with average teachers, average constraints, and students who may not be motivated in the first place. A model that only works when the teacher is unusually gifted is probably more of an inspiring example than a scalable system.