As any software developer will tell you, one great programmer is easily worth ten average ones. The great strides in knowledge have most often come from those we label "genius." Newton, Gauss, Einstein, Feyneman, de Morgan, Crick all seemed to be able to make connections or see patterns that others had ignored. They often visualized the world differently, or with fewer constraints than most of us have on our imagination. There are many great problems of science and society to be solved, and applying genius to them could help speed the solutions.
Perhaps the analysis of Einstein's brain done by Professor Diamond at Berkeley, which seems to show differences in structure in the inferior parietal region, and a higher proportion of glial cells can lead to some physiological answers. Perhaps there are chemical enhancers which can be used (legally, one would hope), to increase oxygen flow to neurons. Perhaps behavioral conditioning when we're young can help create more of the right type of structures, just as musicians who being training in early childhood have larger portions of the brain devoted to their skills.
Whatever the answer, mankind might be better for some more genius directed at the environmental, social and scientific fields.