Carbohydrates can be a tricky nutrient to analyze. Part of the problem is that there are different effects over different time periods. For example, I wrote earlier about how people in the US consumed an average of 500 grams of carbohydrates per day in 1909. So why was obesity and diabetes rare back then? I think there are a few reasons:
First, people were more physically active and therefore they could burn off the glucose consumed (for example, manual labor jobs)
Second, the carbohydrates were generally not refined
Third, people were born from healthier mothers, and began life with normal insulin levels
This third point has becoming increasingly important in recent years. The Paleo Diet newsletter pointed out a new paper that shows how insulin resistance can start in the womb, and then greatly increase the odds of the child being insulin resistant throughout his or her life.
What's happening now, I think, is that there is a whole constellation of factors that are making people less tolerant of carbohydrates. Activity is down. Carbohydrates are refined. And the insulin resistance of one generation is being passed to the next, with each round making the problem worse.
That's why you can look at a population like the Kitavans, and see no problems from a high-carbohydrate diet. They have historically eaten unrefined carbs, and therefore they never developed any insulin resistance to pass down to the next generation. If you feed this high-carb diet (even if it's unrefined) to a modern insulin resistant child or adult, it may end up making things worse.