The Least Publishable Unit Strikes Again
When I was at university, Dave Gifford introduced me to the concept of a "Least Publishable Unit," and explained how in academia, publication was the name of the game. He referenced this in respect to some absolutely brilliant paper by Butler Lampson, noting that the number of LPUs in the paper was almost off the charts and that someone less secure in their position could have strung out the ideas in the paper through a dozen different publications. I would fully expect that as smart people get better at gaming the system, the average content of a paper or patent hyperbolically approaches the LPU, and so each individual paper/patent is almost insignificant, but in aggregate, something more interesting emerges.