My best take on the thing is that, historically, most great physics discoveries were made by generalist, wide-branching natural philosophers. Granted, "natural philosophy" is arguably the direct ancestor of physics from which spawned the bastards of "chemistry" and "biology", but even regardless, the key point is that they were generalists and that, if we were going to solve the current problem simply by throwing more specialized physicists and gamma ray guns at it, this is not the evidence I'd expect to see.

Given historical base rates of generalists vs specialists in physics, and the ratio of Great Discoveries made by the former rather than the latter, it feels as if generalists have a net advantage in "consolidating" recent research into a Great Discovery.

I do have to agree, though, that all of them came from physicists, if not necessarily formally trained, although in most cases they were. Good knowledge of physics is necessary, that I won't argue. But what I'll point out is that I've personally met many more game developers and programmers with a much better grasp of (basic) physics (i.e. first volume of Feynman's Lectures) than college physics department members, on a purely absolute count. It doesn't seem that far-fetched, to me, to assume there's a comparable difference in base rates of people within and outside physics departments with a solid enough grasp of physics for the Next Great Discovery, whatever that threshold may be (and obviously, the lower the actual threshold, the more likely it is that it will come from outside Physics Departments).