Intellectuals get great motivation from big grand questions, and often do work adjacent to them, intending to be relevant to them. Such intellectuals mostly get prestige for the observable impressive detail of their work. Detail such as data collection, stat analysis, game theory models, agent-based models, case studies, phil analysis, engaging public essays, compelling keynote talks, and managing related orgs.
When the world changes its mind on such questions, intellectuals who are sufficiently prestigious due to their details often gain even more prestige from the fact that they had supported those new positions in their work well before the world did. And such intellectuals like to credit their details for contributing to this intellectual progress.
However, intellectual opinions on big grand questions are usually driven more than they like to admit by simple abstract arguments, often processed intuitively. We usually form such opinions soon after hearing of such questions, and then may change our minds as new such arguments occur to us.
When forming opinions based on abstract arguments, and opaque intuitions, it is possible to be either sloppy and fast, or careful and slow. Yet it is hard to tell from the form of the results just how much work went into them. The space of abstract arguments is so vast that both approaches result in pretty simple-looking abstract arguments. As a result, it is hard to impress audiences with such arguments. Which is why prestige is given mostly for associated impressive detail.
The problem here is that even though abstract arguments are in fact typically relevant and potent for drawing conclusions regarding big grand questions, intellectuals spend most of their efforts on less potent adjacent details. If we would instead think harder on the abstract arguments, we’d plausibly make much faster intellectual progress.
Well-subsidized prediction markets re future intellectual opinions on big grand questions would be one way give stronger incentives to think harder about abstract arguments. But don’t expect the intellectuals now gaining prestige via their impressive details to offer much support for this change.
The coming infusion of AI into every corner of academic inquiry will likely devalue the prestige historically attached to marshaling a large fact base to carefully argue a position, since AIs can help so much there. (In math it's the ability to craft a complex proof.)
How this affects the academic reputation and rewards systems built up over the centuries will be interesting. Maybe something like a prediction market that measures outcomes more directly will play a role.
I suspect the asymmetry is even stronger than you suggest: the apparent simplicity of an abstract argument may be almost inversely related to the difficulty of discovering it. A complicated argument advertises the labor invested in it; a successful simplifying insight can erase the intellectual scaffolding required to reach it. But detail has another important function besides prestige: it can expose where an elegant abstraction fails when confronted with reality. Perhaps the productive division of labor is therefore abstract reasoning to identify the underlying structure, followed by detailed work designed aggressively to falsify it.