Sornette's book is quite an achievement both in quantity and
quality. The presentation remains informal and quite readable; it reads like a physics textbook, not a math textbook. The references are very extensive (a total of 832! altogether) and they are a very valuable component of the book. In fact much of the book is about the reference material. You might choose to read the book instead of the 832 references... I think this is
There's probably nothing wrong with this book besides the fact that it throws it all at you at a high degree of sophistication and in as terse a way as possible, it seems. It's a unique and beautiful achievement but because it is so dense with information and insight, it seems every word counts for ten and you'll want to read several chapters again and again. Also, even though there is a clear unifying theme from chapter to chapter, the book simply ends almost like in the middle of a sentence. After machinegunning out 392 pages of material at research level spanning quite a few scientific fields, there is absolutely no attempt to put it all together. It's up to you to do it and it almost seems like the author is indirectly suggesting you start reading it all over again to "get it"... So, for the second edition, perhaps the author will be bold and add ten pages of wrap up material at the end so that this will read less like an atlas. Apart from that, it's the best!