The gist of the book is Big Data provides unmatched opportunities for social science. Particulary, Big Data allows us to zoom in to test hypothesis that we won't otherwise be able to. The author provides the examples of A/B test.
What makes me uncomfortable is the author's choice of Google data as honest data while Facebook data as biased data. It looks to me the author jumps directly to the conclusions without necessary proof. To me, how to identify the integrity of the Big Data is a missing chapter in the book.