Thanks for the interesting article Jay Lynch.
Your article refers to the situation where students (and many educators) confuse errorless work for learning. And the reverse is true — when students need to think they often believe they aren’t learning. In my current work I see this frequently.
I agree with your basic premise of using forgetting as an effective learning strategy. However have you tried your suggested approach ‘in the wild’ with students?
While the science of these methods is sound in isolation often they don’t work in real life. By their nature these studies tell you that if A happens then you can expect B. But you can’t force students to do activities. And if they aren’t practicing then it doesn’t matter how good you design the practice to be.
I run a tutoring business and am building adaptive software that helps our tutors to best support their students. We are building in many features including adapting question difficulty to ensure students are being challenged. However it has been hard to convince students and parents that some struggle is good for learning. Using the software students now receive more challenging questions (through a mixture of temporal and other factors) which results in more errors. Our data suggests that if students get 3–4 questions wrong in a short space of time they will simply give up on the homework. And this is a negative outcome because now they aren’t even lifting the light dumbbells as you mention.
You suggest that one can overcome the motivation issue by encouraging students to value brain burn. I am sceptical that this would be effective. Here what you propose is something that we (as the educators) want to happen but not what the student wants. The basics of good design is understanding that people have their own ‘jobs to be done’ (refer to Clay Christensen’s disruption theory) and you can’t change that, at least not quickly. Almost all health and fitness products based on good research have failed because they relied on behaviour change. Instead the ones making money are those lose-weight-quick schemes that fail to create lasting change. The educational solutions that work will need to account for the needs of the students as they see them, and not what others (parents, teachers, product creators) want.
Therefore I see a trade-off between the efficacy of approaches in pure form and how long they work due to motivation. It then becomes an optimisation problem where you’ll want a solution that balances effectiveness and motivation to maximise learning overall.