I agree with Mark that James Wood's condescending comments about Palin's use of verbage are pure de-haut-en-bushwa. On the other hand, let's not delude ourselves about this item. Palin's verbage is not simply a term for "language" or "wording" that has been happily circulating in vernacular speech since it was first attested 200 years ago, in defiance of the assaults of prescriptivists. Verbage is not colloquial English — I mean, people don't go around saying, "Hey, Sparky — watch your goddamn verbage!" It arises as an approximation of a fancy-pants word that people have seen in print: it's a lot more plausible to assume that people would misread verbiage as verbage than that they would mishear it that way, particularly since this is a re-analysis favored by analogy. The fact is that in both its form and its meaning, verbiage is a weirder word than most people — including the editors of the OED — realize.
One reason why the reanalysis of verbiage as verbage is so easy is that it isn't obvious what that i is doing there in the first place. The root verb- yields words like verbal, verbose, not verbial or verbiose, and there's no reason why the suffix -age should introduce an i (the i in foliage comes from the Latin root folium). Then too, why should the combination of verb- + -age imply prolixity? By all rights, the word should mean simply — and only — "wording," as Palin assumes it does. (After all, acreage doesn't mean "an excessive amount of land.")
As it happens, though, the word almost certainly doesn't come originally from verb- + -age, as the OED says it does. According to Alain Rey's Dictionnaire Historique de la Langue Française, the word verbiage first appeared in French in 1674. It was derived from the the Middle French verb verbier or verboier, which he glosses as "gazouiller" ("chirp, warble") as applied to birds, and which he connects to a 13th-century Picardian verbler, "warble, speak in singsong" this in turn derived from Frankish *werbilon, "whirl, swirl" (cf German Wirbel, "whirl") "Since the 17th century," Rey writes, "the derived form verbiage has been connected by popular etymology to verbe and for its meaning to verbeux [ = 'verbose, wordy']." Since the English verbiage isn't attested until 1721 (and then, as it happens, in Prior's poem about Locke and Montaigne), it seems quite likely that it was borrowed from the French — this would explain both its idiosyncratic form and its unaccountably restricted original sense. Both morphological and semantic analogy, then, would favor a popular reanalysis of the form of the word as verbage and of its meaning as "wording."
But whatever the derivation, my guess is that the majority of literate speakers consider the verbage pronunciation an error or an illiteracy, and that a smaller but significant percentage of them would also say that it was a mistake to use verbiage to mean simply "wording" without any implication of wordiness, even though as the OED indicates, the word has been used in that sense by reputable writers since the early 19th century. It's too bad the American Heritage didn't record that opinion or ask the members of its Usage Panel what they thought about that variant, in the interest of alerting its readers to the potential risks in using it (it isn't only the James Woods of the world who consider it an error, after all). I take some responsibility for the omission, since I was chair of the Usage Panel when the entry for this word was prepared. But it's in the nature of the lexicographical enterprise that you can't get around to everything, and silence shouldn't always be equated with unqualified authorization.