This may have been true a hundred years ago, but it hardly seems true of academia right now. Maybe you're thinking of some specific field with these characteristics, and broadly generalizing?

I should have been clearer that the claim was that industry cares more about results, less about novelty, and less about individual effort than how much academia cares about those three, not that academia strictly prefers novelty to results. I'm aware that the average number of authors per paper is increasing, and have been involved in projects with ~10 PhD physicists (plus many graduate and undergraduate students) all collaborating on experiments run on a single device, and know that when you look at astronomy or plasma physics or particle physics you find hundreds if not thousands of people working on huge experiments--but I'm under the impression that those are more government projects than they are academic projects.

The comparison I'm making is between, say, three graduate students who work together on a project and three cofounders who work together on a company. In the first group, each person is expected to produce and defend work that will stand alone with just their name on it; in the second group, each person succeeds or fails with the group. I'm not aware of any set of multiple graduate students all working together on the same project, and get the impression that many (if not most) software startups have multiple founders these days.

The Polymath project is exactly an example of some super-mathematicians crowd-sourcing the implementation details of their great ideas....

As I understood Stefan_Schubert's suggestion, we would have external agents (like grant agencies) identifying the super-scientists to think great thoughts, and the journeymen to implement them. My claim was that the external agents, and the classification system, were not necessary, but I recognize that was not worded as obviously as it could be.