I wish the hundreds (or even thousands) of hours from all those people wouldn't be wasted solving the same didactically useful but rather useless results of manual work. Why can't the manual labour put in result into something great or useful like a working program, framework, bugfix, codechallenge (I surely am fantasizing a bit here but you get the idea)? That was always something that bugged me about doing CS as an undergrad (and grad): all those hours of brainpower are wasted solving the same problem instead of looking at the final result from all the lessons and picking the best out of each assignment and having that at least run against last year's class. Education surely is ripe for disruption and this would be one way to make use of the larger scale of having a larger number of people work on problem sets.