I agree with the risks, but it seems like the solution, at least at the undergrad level -more, and more effective, in-class evaluation, wouldn’t be that hard to achieve. Classes centered around the term paper are the most challenging, but how many of them need to be?