Did they hire all of these people just because they could? Are the projects that inefficient? Or have they killed future growth for Wall Street?
Anecdotally, yes. In my experience most projects are carried by a select few 'Heroes' doing the majority of the difficult implementation and innovation, tied together by a larger volume of less impressive but still critical connective work.
From my reading, the author's suggestion is that, functionally, most teams understand their goal and how to work towards it without regard or 'real' benefit from an overseer.
It's a strong claim, and one that should be relatively easy to test if an organization or research team wanted to.
But that's the nature of any project. 80% of it is 20% of the work. But the remaining 20% - the wicked problems - takes up the remaining 80%. Any given project, by "law", is going to require a hero or three.
Put another way, if you had the heros doing the less heroic stuff not only would that be a waste of resources but eventually they would update their CV and leave.
They are, in their argument, almost entirely composed of "non-heroes" and their actual work is of speculative value. Hence why they are suggesting that 50% of them can be randomly removed without a corresponding loss of effective productivity.
I don't know if that's true or not (though I'd love to see it studied), but I can at least see that logic behind the claim - roughly summarized as:
"The ability of a manager to increase the productivity of workers is inversely proportional to the complexity of the work being done."