The NSF budgets millions to fight a problem not demonstrated to exist
whyevolutionistrue.com
whyevolutionistrue.com
The word "probably" is doing immense heavy lifting in that sentence.
Still, it doesn’t make the crux of the author’s argument any less sound; if you want to invest a huge pile of money into solving a problem, you’d probably want to be sure that the problem is well-defined and actually exist.
The "scientific method" doesn't tell you what to study, or how to go about studying it. It's ultimately a matter of intuition about what is most likely to be productive. Your intuition is guided by your experience, but it rarely gives you a "this is what you must study next" pointer.
Funding is yet one more layer on top of that, since it has to balance expected financial returns to society with things that aren't ever expected to make money. Sometimes they do anyway, but that's not why we do them. We do them out of a sense of curiosity -- the same reason we make art.
I suspect that on balance our "just plain curiosity" work has paid for itself, but only because a tiny number of things (like quantum physics and number theory) have had insane returns. The vast majority will not, and you cannot tell which it will be.
Your phrase "political logic" has all kinds of negative connotations. But ultimately, politics is the process by people work together, and science is one aspect of people working together. That's messy; it's always messy. But it's also a real thing, and we cannot pretend it doesn't exist just because we can't write equations for it. Because if we set that standard, we'd have to abandon science itself.