From deadcities_icon a discussion of reviewing (though, for some reason, online reviewers are being accused of giving craptastic reviews while print gets a pass).
My take: It seems (no offense intended, gabe) that this post ignores the definition of literary criticism and bulls on ahead by substituting an undefined term in its place. Criticism and reviews are completely different animals. A review that included an exegesis of the Marxist allegory in a book's subtext would definitely pique my interest, but only from an academic standpoint. I'd read the book, surely, and I'd demolish the piece of criticism if I could, but here's the rub: a review that attempted to do that with even the shortest of novels would be -- at a minimum -- 30+ pages. Who the fuck's gonna read a 30 page review?
Criticism attempts to codify literature in an OBJECTIVE fashion. It's a pseudoscience, in that it wants very badly to treat every work to the same examination under the same conditions. In this way, much critical literary theory attempts to create a paradigm in which aspects of literature validate the critic's worldview. A cultural materialist from the Frankfurt school is going to find examples of how the book demonstrates the top-down model of societal evolution. A semiotician, or psychoanalytical linguist, on the other hand, is going to examine, among other things, examples of the attempt to depict a unified self and the resultant, inevitable cycle of fracture and rapproachment of either a character or the author; some even psychoanalyze the reader-response.
Reviews, however, are much, much more basic an animal, and far more interesting a read for the casual consumer who simply wants some purchasing guidance. While the reviewer's job requires (sometimes) an approach less journalistic than creative, his job is to impart some very basic, SUBJECTIVE information: What kind of book is it: What kind of writer wrote it: How much or how little he liked it. That's really the crux of it, an unabashed OPINION is the aim of a decent reviewer. If a reviewer applies consistent standards to every review he gives, even if his audience disagrees with him, he's golden. He says he hates a book, he says he likes a book, that matters only insofar as he's honest in his estimation and thorough in his explanation. Anyone who expects more from "the man of taste" is either overdemanding or naive.
To sum it up, criticism is (nominally) objective, scientific, repetitive, analytical. Reviews are subjective, creative, always fallacious, hopefully interesting and entertaining. Different beasties, no? Uh, duh, says I.