Most of my research time this semester has been spent on educational research. It's interesting, as I've started to see the same pitfalls in this literature as the nutrition literature.
A key problem many people propose simple, binary solutions to issues. In nutrition, some people have fallen under the spell that high-carb is bad, and low-carb is good. Regardless that it's not accurate, it produces a nice, simple mental package that requires no further thought.
In education, you have a similar disagreement between the "constructivists" and the lecture group. The lecture folks argue that students need direct instruction. The constructivists argue that students should be allowed to explore and be guided to create their own knowledge.
In actuality, a combination of the two is best. Studies have shown that having the instructor present a worked example and then having students try problems is superior to either giving all examples or all problems. Unfortunately, this solution doesn't give anyone that neat, binary answer they were looking for.
I wonder if the tendency towards binary solutions is a part of our evolutionary psychology. Maybe it's existence allows mental energy to be saved. Unfortunately, its potential existence also biases people towards simplistic, one-dimensional solutions.