One thought though: is it really just genius, and the only way to get to that top right compelling and unique state by just jumping there?
Feels like it would be more of a combination: someone tries a unique idea and it's not compelling. Someone else sees this idea, but changes something about it, so it's now unique and slightly compelling. This gets repeated until you end up with something unique and compelling. This new idea now gets copied exactly, and it slowly moves from the top right to the bottom right - compelling but non-unique.
ie. Is this not a continuous process curve innate in all invention? The only real effect might be the speed of movement along the curve - someone (a genius) may be able to iterate across the curve by himself by building on his own ideas?
If this is true, then genius is not really required - all that we require is constant iteration of bad unique ideas until they become something good. Seems to hold up in reality - the more people involved in something, the quicker it gets somewhere good, as there are more iterations happening.