A radical thought: We (educational researchers and instructors) spend great time and energy trying to optimally engineer our students’ learning environments and experiences — pacing, sequencing, balance of examples vs. tasks vs. information, cognitive load, collaborative designs, testing intervals, reward structures — drawing on a great deal of disparate research, collective and personal experience, and intuition.
Perhaps the human organism is well-adapted enough that if we can give learners the freedom to pursue their own learning, with an adequately rich and suitably organized and accessible array of resources, they would naturally find a highly optimal balance of these factors?
In other words, maybe people know how to learn better than we know how to teach, and we just have to figure out how to let them do it. And stop convincing them that learning something like Physics means doing it “the school-like way.”
Afterthought: I suspect we can’t do this because so much of the educational enterprise is designed to get learners to learn things they’re not particularly invested in learning. Maybe we need to revisit that?