I’ll leave those who know Powers’s work better than I to judge the validity of Deresiewicz’s criticism with respect to Powers, but I do agree with this as a general statement on literature:
But intellect and scientific acumen are not synonymous, though our culture seems to thinks so. "It’s not rocket science," we say, or, "It’s not brain surgery." So a novelist who understands science must be really smart, and a really smart novelist must be a really good one. (Was Hemingway "smart" in this sense? Was Austen, or Proust?) This confusion is no doubt compounded by the fact that, like most people, the typical book reviewer is unfamiliar with, and easily intimidated by, scientific concepts, and thus apt to defer to, if not genuflect before, those conversant with them. It is further compounded by our dimly understood but longstanding desire to "bridge the two cultures" of science and the arts (another phrase that crops up in Powers’s reviews). From Matthew Arnold to C.P. Snow to today, there’s been a vague feeling afloat that if only somehow those two modes of knowledge could be made to talk to each other, science would be humanized (whatever that means) and art made relevant to the scientific age (as if it weren’t already).
I doubt this demand will ever be satisfied, for the simple reason that no one really knows what it means, least of all the people who make it. But certainly one way it won’t be satisfied is by treating the novel as a container for scientific ideas.
I think there’s much to be gained by using scientific conepts in literature. I think they accurately capture certain facets of a time and civilization (and even certain personalities), and I think they can work brilliantly as metaphors. But I agree that they should remain subservient to the development of the work they are included in. For a perfect example of this done well, see Michael Frayn’s play Copenhagen.
For what it’s worth, this review in Slate suggests that Powers doe exactly that.
After reading Powers, C.P. Snow’s once-famous complaint about the "two cultures"—scientists and humanists, each unable to listen to the other—melts away. The novelist trained as a physicist himself. No wonder he gets celebrated as a cerebral novelist, as an explainer, as the smartest writer on the block. Yet the interest in Powers as a man of science misses what keeps his characters alive. All his information-rich protagonists—teachers, programmers, professors, singers, accompanists, homemakers, hostages—have to master a vast array of data: All of them make, from that data, refuges, new spaces, kinds of art. All of them (Powers argues) need both the arts and the sciences in order to share a household, a nation, or a world.