That's well said about the need for the arts and sciences to work together for the benefit of the future. Currently, that's not happening so well, and I can think of a number of reasons why that's so. First, there's scientism, a condescending prejudice against the arts for being insufficiently rigorous.

That may be just a rationalization for the more fundamental, second reason, which is capitalism: science and engineering are worth more in capitalistic terms because it's much harder to profit from the arts and humanities.

And that economic reason is exacerbated by a third, political one: democracies are vulnerable to demagogues who thrive on regressing culture, on dumbing things down so their cults go unnoticed. (Social media and high-tech accelerate this infantilization too.) Culturally, then, the masses don't respect the arts because they've become anti-intellectual. We consumers think we're experts on everything because we've fallen for the advertiser's hype and flattery. We're know-it-alls who think we should be getting art for free on the internet. The creator economy is thus neofeudal.

Regarding your first point, though, about the levels of analysis, my question would be whether you think the higher levels reduce to the lower ones. If not, we can posit the existence of the simplified properties on at least pragmatic grounds, in which case the "illusion" of central command is real enough for explanatory purposes.