This recently published PNAS article constitutes the first experimental evidence of Wrangham's hypothesis concerning the importance of cooking in human brain evolution. The article notes that our brains consume about 20% of our body's metabolic energy, compared to 9% in other primates, raising the question as to how we could afford to feed such metabolically expensive organs. The paper argues cooking bridged this gap, following several lines of evidence, including: 1) brain size is directly linked to the number of neurons in the brain and 2) the number of neurons is directly correlated to the amount of energy (or calories) needed to feed a brain. After adjusting for body mass, they calculated how many hours per day it would take for various primates to eat enough calories for uncooked food to fuel their brains: 8.8 hours for gorillas, 7.8 hours for orangutans, 7.3 hours for chimps, and a whopping 9.3 hours for Homo sapiens. In sum, the paper seems to make a relatively good case that cooking lifted an important constraint on human brain evolution. However, what seems much less clear, based on these results is whether cooking itself was responsible for the first dramatic burst of brain growth and how, by itself, may or may not have driven this expansion. Thus, Wrangham's more modest claim seems supported, while his more extreme claim does not.
Here is a link to the article: http://www.pnas.org/content/early/2012/10/17/1206390109.full.pdf+html
As well as an interesting Science news blurb summarizing most of the important points: http://news.sciencemag.org/sciencenow/2012/10/raw-food-not-enough-to-feed-big-.html