This book has much good information if you know exactly what you are looking for. However, its tone is very dry and academic and lacks a big picture perspective. It glosses over essential information on mechanics and in just 200 pages covers Newtonian and Lagrangian concepts. There are lots of equations, but the examples all use variables, thus making the examples no clearer than the equations they are supposed to be illustrating. There are no numerical examples at all.
Chapter 5, on physics engines, which the author himself says is the heart of the book, is woefully inadequate. He talks about the equations and features he wants to implement, and shows code snippets for the features, but there is one essential component that is missing - block diagrams and accompanying discussions that illustrate how all of these pieces fit together. A crate full of clearly labeled mechanical parts does not an automobile make.
Next we come to one of my areas of interest and one of the main reasons I bought this book in the first place - Chapter 6, "Physics and Shader Programs". The author certainly hits some fascinating topics - ocean waves, skin and bones animation, etc. However, once again the author skips what is between the basic physics equations that comprise these effects and the implementation of these equations and effects in code. The code that he does show is in NVidia's Cg language. So if you don't know Cg, the code is undecipherable. I would have preferred algorithmic pseudocode. Then I could implement the effects in the language of my choice.
The last four chapters of the book take an uptick in quality. These chapters are strictly on numerical methods, and I can't believe they are written by the same author that wrote the first six chapters. There are fairly good discussions and examples on the subjects of linear and mathematical programming, ordinary differential equations, numerical methods, and quaternions, and also how these mathematical methods fit into solving problems in physics. They are not self-contained tutorials, but they are a good summary and review for the physics student who has already had some exposure.
I really can't recommend this book as anything but a reference. Its merits are mainly the tying of mathematical methods to physics and good basic discussions on some topics. But where the rubber meets the road - answering the questions of "How does one go about building a physics engine?" and "How do I implement physics within the confines of a computer game?" I think that this book is a failure. I give it three stars only because the author obviously put a lot of effort into the work, and what he does write is not incorrect.
If you are interested in game physics I recommend "Physics for Game Programmers" by Grant Palmer. It takes an engineer's viewpoint of the subject and features plenty of examples in Java of how to simulate explosions, lasers, and the motion of various classes of vehicles. There is no discussion of game engines, though. A good reference on college level physics can be found in the most recent edition of Schaum's Outline of College Physics.