I think people mostly waste time at the universities learning things they will forget. Giving students a novel problem to solve and ask them to come back when they find the solution might work better in terms of sticking knowledge. This doesn't necessarily need to be a group project.
It would be great for learning, but requires an enormous amount of work for the academics involved. That side is pretty important in explaining why the system is the way it is. It is due to have some novel ideas injected about how to balance effectiveness and efficiency.