If you're a designer, you haven't truly lived until a classroom full of students brings you paper prototypes of their video game proposals to evaluate. This was the most fun I've had grading. Ever.
Pretty much every student had at least one thing in their prototype that the rest of us (including me) can learn from, even if it was "don't do this". Failures of prototypes usually teach more than successes. Some students (correctly) failed on their own several times, threw away their own work and brought in something that worked much better -- allowing them to discuss failures AND successes.
I did have some assignments like this back in the Fall, but the students were leaping right into prototypes without first writing up a formal treatment (and also without having a solid foundation of game design theory to build on), so I think for best results this really should be a three-step process (theory, then treatment, then prototype).
The only part I had difficulty with was when a student brought in something that served as a negative example. Having your prototype harshly critiqued in front of an entire class would be mortifying, so part of me wanted to just move on without too much discussion... but on the other hand, part of me wanted to call this out as a great learning opportunity for the entire class. I think I managed to split the difference, embarassing a few students without providing the education for everyone else. Some day with more practice I'll find a way to do that better.