I was thrilled when I learned about this book and I waited impatiently for it to arrive from Amazon. Boy, was I disappointed!
The idea of such a book is great, somebody should have done it. The execution though is the one that is bad. The book is mostly focused around small number of defining concepts, which are supposed to explain and substantiate all the facts about the way brain works and the suggestions of how to become more efficient in whatever you do. These concepts are the L-mode and R-mode of the brain, the Dreyfus model of skill acquisition, and the metaphoric comparison of a brain with a two-CPU computer.
Unfortunately, L/R-mode theory is now considered wrong and dated (the theory is more than 20 years old -- a lot has happened in neuroscience since then), and basing and substantiation suggestions on it is questionable. Even though the suggestions themselves are mostly reasonable and useful (in case you have not come up with them on your own yet), the constant L/R-mode preaching makes an impression of somebody selling you snake oil. The L/R-mode explanations make up a bulk of the book, sound really fishy, and get annoying pretty quickly.
Dreyfus model, although somewhat useful in some fields, not too useful in the context of research work and science (and any non-trivial software engineering), where things are a tad more complicated [note: this is my personal opinion, don't take my word on it and read about it elsewhere if you want]. That wouldn't be a problem, if Dreyfus model wasn't used throughout the book to explain things.
Comparing a brain with a two-CPU computer is just blatantly wrong, the way the brain works is not even in vicinity of how CPUs (and the related wiring) work -- just read some other books and research papers on the subject. Thus using the metaphor abundantly in a book which tries to give an impression of a book where the facts are checked and substantiated is questionable.
Of course, that's not all. I found many places in the text where something was stated (which wasn't obviously true or false), but as if it was following from some other facts. If you're not careful enough when reading, you are likely to learn something that isn't.
Less important things which I didn't like: the narration and the design/formatting/images and text relevance. From the start the author notes that this book is not necessarily intended for programmers, however the text is full of irrelevant programming allusions which would bore any non-programmer to death, without any chance of getting any useful meaning from the allusion. Heck, I'm a programmer and I was bored and struck with superfluity of these examples. Oh, and don't forget about smileys in the text. Don't get me wrong, I'm not narrow-minded, however I still believe that well-edited text in a book on a serious topic could do without smileys and still be able to communicate jocular mood if there's need for it. The book is full of irrelevant examples and images (I love images, provided they are useful!), take the "unix wizard" image as an example. There's even an awful attempt at infographics (p.229, fig.8.4, "Relative IQ point loss") which takes almost half a page and is really a bad example of using a bar chart.
Another annoying thing is that another book by the same author - "Pragmatic Programmer" - is praised persistently throughout the text. Although it's not a bad book, there should be some restraint in self-advertising.
To be fair, there are some good suggestions and practices. It's rather unfortunate that they get diluted by a mush of bad science and lacking narration.
Verdict: if you want to spend your time reading a good book on the topic, go read something else, for example, Medina's Brain Rules: 12 Principles for Surviving and Thriving at Work, Home, and School, Weinberg's Becoming a Technical Leader: An Organic Problem-Solving Approach, or DeMarco's Peopleware: Productive Projects and Teams (Second Edition).
