I most value this book for its brief and lucid introductions to the physics (e.g. multiphase flows) that can be incorporated into lattice Boltzmann methods.
As far as the lattice Boltzmann method is concerned, I soon found myself relying on papers in literature rather than this book (although it has to be said that this book thoroughly references publications in literature).
My main criticisms arise from following points:
* as a quick start guide, this book falls short in not discussing issues like initialisation and the general streaming-collision framework. I found the latter particularly confusing, since their code snippets suggest that they solve a slightly different equation than the one analytically described in the book. Moreover, the extension from 2D to 3D (which is left to the reader) is not as trivial as the book suggests (e.g. for the boundary conditions).
* the book gives an introduction to phenomenological models for e.g. multiphase flows. If you are an engineer or geoscientist, you might want to use models that are quantitatively more correct.
* The book doesn't go beyond lattice-Boltzmann toy models. If you want to do something "real" with lattice-Boltzmann, you will need to address more advanced issues (like how to deal with curved boundaries, or with higher-order lattices). While one cannot expect from the scope of this book to address those issues directly, it is a pity that the book doesn't prepare in any way for those issues. So probably, you'll have to throw away your code and start all over again if you ever want to model something more.
* Due to ongoing research, this book starts to become outdated. In part because the book introduces lattice-Boltzmann models from lattice-gas cellular automata (as they evolved historically), instead of being directly based on the Boltzmann transport equation (as is more common nowadays).
Overall, I quickly abandoned this book while writing my lattice-Boltzmann code. Nevertheless, you can probably have a lot of fun with the models proposed in the book, as long as you don't want to do anything too useful with it.