Spiral learning: it's like your toes.
MIT professor ever).
Undergraduate life here at to which she contributed, "They (the MIT students) are used to a certain kind of learning pain at an enormously high dosage level; but you're administering a different kind of learning pain that although less, actually hurts more." The light went on in my head. If you're used to taking it in the face all the time, you're ready for another; but when you get your toes stepped on for the first time it's going to hurt because the pain is brand new.
I remember balking at the lack of direction in ECS my freshman year (To this day I am not sure how I passed) and privately whining about the vagueness of UOCD (user-centered design) at the start of the semester sophomore year. Then I started becoming a more proactive TA and talking to the profs about why they were causing us pain. And I started realizing they weren't trying to teach us math per se (or circuits, or whatever), but a different way of thinking.
Ok, that sounded corny and vague. But it's true; Olin students hear ad nauseum how we'll forget all the formulae but remember how to learn things in order to solve problems. The thing is that in order to teach methods (ways of thinking), you need to practice applying them to stuff, and in our case that stuff happens to be engineering content, like thermo.
Learning the many different ways to "think like an engineer" is infinitely harder than learning how to apply the Fourier transform for exactly the reason Maeda outlined: every different way of thinking you learn is like being attacked in a new and painful way in a new and painful spot. Used to getting your toes stepped on? We'll start pelting eggs at your kneecaps. You learn how to adapt, rather than how to adapt to a specific situation. You become a generalist, a designer, a problem solver.
In short, an engineer.