On pretty much every subject, there's a lot of crap out there. Some people are better at marketing what they know than at figuring out if what they know is right, and this leads to a lot of misinformation. Parenting is no exception, as you can find "experts" on opposite sides of almost any topic. In The Science of Parenting, Margot Sunderland aims to cut through the crap by giving parents advice based on science instead of conjecture.
For the most part, I think she succeeds. On a number of topics, including sleep training, how to say no, how to play etc, Sunderland references a great deal of research including long-term behavioural studies and our latest knowledge of the brain.
For the most part, her explanations for her suggestions sounded pretty plausible. But that worries me a little. I don't know how well she vetted the studies she referenced, because as we know, a lot of study results cannot be replicated because they are based on bad science.
Sunderland didn't really debunk anything, which makes me think she just took a whole bunch of stuff at face value, which probably has resulted in a lot of inaccuracies. For instance, the section on nutrition, a topic on which I have read a lot of credible stuff which debunks the mainstream view, was a little too naive in my opinion.
Nevertheless, basing one's parenting decisions on attempts at science is still probably better than nothing. As such, this book is probably better than 90% of what's out there on this topic.