I can't quite put my finger on it at a glance, but my immediate instinct is that this sounds nice at a very superficial / philosophical level, but actually flie

I can't quite put my finger on it at a glance, but my immediate instinct is that this sounds nice at a very superficial / philosophical level, but actually flies in the face of actual pedagogical literature over the last few decades. For one, you're describing research, not education. The fact that it "can" be done (sub-optimally) in this manner, and the "this is how we've always done it" impetus behind it, doesn't necessarily make it a good idea educationally.