The Shangri-La diet and Atkins diets would be in that category, as would the entire contents of Four Hour Body by Tim Ferriss.

I disagree with respect to the Four Hour Body. While Ferriss is particularly good at presenting his content to a popular science audience, using enthusiasm and anecdotes to make the information salient, the scientific basis in the background isn't bad and definitely not in the same category as Shangri-La or Atkins. (I refer here to the presentation in his recent book, not the even more hypeish blog post from 2007 with the rather provocative and unrealistic link-bait title!)

What is particularly notable, and unusual for this kind of health and fitness book is the entire chapter on "Spotting Bad Science 101". It opens with a section on the difference between correlation and causation, a principle that is sprinkled throughout the rest of the book in notes and reminders. It would clearly be unfair to apply the criticism "people who assume correlation equals causation are often not the best at medical literature search and summary" in this case. Particularly if extended to the 'entire contents' of the book!

A criticism that I would level at Ferriss is that he takes weight loss principles that are fairly well known (in some circles) and repackages them under his own brand name, complete with the 'faddishness' that you mention. Plenty of blatant attention seeking all round.

I don't endorse everything in The Four Hour Body; it is not a textbook. It's just the best book that happens to be out there on the subject that is accessible to the intended audience. It models an instrumentally rational approach more so than a purely scientific one, balancing the use of established science with trying to incorporating other forms of legitimate but less rigorous forms of evidence from evidently successful professional advisors and simple experiment by hackers.