Having not seen the MS, I can't comment on whether it was a good paper or not. Mind you, the speed with which Nature processed it suggested that they probably didn't read it either. It's questionable whether it would have made any difference if they had. Although, in a recent chest-beating defence of its editorial policy, the journal claims that "our editors spend several weeks a year in scientific meetings and labs, and are constantly reading the literature," this doesn't make them scientists. It just makes them better-informed about science than most other journalists. This is a little worrying when you consider this statement, from the same defense:
"Another long-standing myth is that we allow one negative referee to determine the rejection of a paper. On the contrary, there were several occasions last year when all the referees were underwhelmed by a paper, yet we published it on the basis of our own estimation of its worth."
Sometimes that works well (e.g. Watson & Crick's 1953 paper on the structure of DNA, which Nature published without peer review) and other times it doesn't (Jacques Benveniste's 1987 paper "proving" homeopathy).
The error that people make when they get all aerated about Nature's seemingly high-handed policies is mistaking it for a scientific journal. Truthfully, it isn't one, or not as most scientists would understand the term. It's a news magazine for a scientific audience. It forms part of a PR continuum for research that runs from Nature at one end to a brief paragraph in the local paper at the other.
In my own field, for example, Nature is used for the initial announcement of important fossil discoveries, but the publication that has the longer shelf life is invariably the monographic description of the specimen. This usually arrives some months or years later after careful study has been completed. It is comprehensive, detailed, has benefitted from proper peer review, and is often repeatedly referred to and cited for decades afterwards. By contrast, a "Letter to Nature" is the scientific equivalent of yesterday's papers - on the bottom of the budgie cage by the next day.
What struck me as I read Cozzuol's impassioned defense of the rigor of his group's work on the new tapir, encompassing morphology, paleontology, 3 sets of mitochondrial genes, statistics, cladistics, etc., was that here was a study that was wasted on Nature. It needs to be published in some heavyweight scientific journal, where they can lay out their evidence in full, have it read by an informed audience, and (if it stands the test of time) added to the canon of literature on mammalian evolution.
Then (as Cozzuol himself rather tartly pointed out) someone else will get to publish a comments article about it in Nature and help themselves to the impressive citation impact factor.