Kent's essay, noted below, links prosody to what was perhaps the main divide in 20th-cent. American poetries - bigger than that between "raw vs. cooked" or "New American vs. New Critical" (at least before the advent of deconstruction and langpo, anyway) :  the divide between symbolist/"idealist" and objectivist/"realist"  (Stevens/Crane/Eliot vs. Williams/Pound/Zukofsky, for example).
 
The divide represents a sort of inherent conundrum of thought & perception.  B.J. Leggett's book Wallace Stevens and Poetic Theory argues that Stevens actually played both sides, pondering & negotiating his way around issues of imagination & reality.  His skill in doing so registered in the fascination, controversy, sheer volume of criticism of same.
 
I'm always jealous of this kind of well-organized well-informed essay.  But leaning as I do toward the Stevens end of the spectrum, I guess, I have a few quibbles.  There's a value judgement being made:  the line which includes Rakosi is supposed to represent a humbler, more active/passive, more direct apprehension of "things as they is"; Stevens' mode is seen as imposing a priori categories of interpretation on experience.
 
A problem with this analysis arises, though, when you grant the possibility that reality itself and the things of reality display order, logos, Dike, Tao.  Then the self-reflexive and circling "prosodies" - the rhythmic patterns which Kent describes so succinctly - may be understood as poetry's function or capability to evoke or represent such a situation of "realities-within-Logos".  And this may be, ultimately, at least as realist as the more indicative particulars of the WC Williams vein.