Michael Pollan's "Unhappy Meals" article for the NY Times magazine was fascinating in its argument against "nutritionism". And here's Daniel Engber's rebuttal in Slate (which largely seems to argue along the lines that Pollan's conclusions seem to smack of the naturalistic fallacy). Although both seem to agree on the"Eat food. Not too much. Mostly plants." conclusion that Pollan comes to.
Meanwhile, badscience answers the question: "why don’t doctors, dietitians, and genuine nutrition academics make the same elaborate claims for the miracle powers of individual foods that journalists, manufacturers and gurus do?" Short answer: pop nutritionism is bad science.