I agree, all the easy problems are solved so only hard ones are left, but the issue of widespread knowledge doesn't seem to be as important to me. The difficulty in a lot of fields is not for example picking up undergraduate knowledge in game theory or whatever, calculus was once an advanced topic but we can teach it to 12 year olds now, but making deep contributions to fields because as you say the easy problems have been picked so either you have to introduce a new paradigm or you have to have an ever growing amount of prerequisite material to know beforehand to contribute something new of significance.