It's only an aside to your main point, but "work on the most important problems you can think of" is a horrible heuristic for everybody to follow. If you want to advance human knowledge, then work on the most important problems you can think of, as weighted by your best estimate of the likely time advantage of "when I would discover a solution" over "when someone else would have discovered a solution". Unless you are one of the smartest people in the world, this weight is likely to be negligibly greater than 0 for many of the most important problems you can think of, simply because there are already plenty of smarter people already working on them.

That's okay. Many of the most important solutions in the world came about because someone was working on an unexpectedly-related problem that would never have made it onto a "most important problems" list.