Taubes has recycled a hypothesis that was discarded in the 1980s. Sadly, the whole weight problem is much more complex -- a mix of cultural changes, prosperity, hyperpalatable food, etc. Google Kevin Hall. He's had a couple of studies in the news recently that nicely refute the insulin hypothesis. His most recent study, which is not yet published, was funded primarily by Gary Taubes' NuSI foundation, too.
If you pick a random chapter of one of Taubes' books and look up his reference, you'll also find that he is quite dishonest or at least confused about his referencing. Nina Teicholz almost completely replicates his referencing in her book too. Plenty of people, like Seth Yoder as an example, have gone through chapter by chapter and dug out references to show that they have been misused. Misusing references is a fine old tradition, popular with alternative medicine enthusiasts like Adele Davis, etc. Some of the misuse is downright dishonesty and some of it is just picking and choosing the bits and pieces that fit the hypothesis to find the pattern that they want to believe is hidden in plain sight. Science doesn't work that way, though.
That doesn't mean that you can't lose weight with a low carb diet because you can! I've seen people lose weight with high carb, low carb, and all kinds of goofy diets in between (Eat Right 4 Your Type!). Most diets work as long as you work them and watch your portions -- and most people, especially women, who are successful with low carb weight loss have to watch their portions too. I tried high carb, but it left me so hungry that I gained weight and obsessed about forbidden foods. Portion control and tracking works much better for me because I'm too skeptical to convince myself that entire classes of food are "bad for me".
The National Weight Control Registry is a fascinating resource for real weight loss and long term maintenance data.
