I bought this (not on amazon) a few weeks ago. I teach statistics and also statistical graphics, so I try to stay abreast of what's on the shelves to recommend to students. I'd tell them to give this book the once over in the library, but not to go straight out and buy it. Statistical graphics have two interdependent aspects to them; how they look, their design and how much information they convey, their content. As I said they are not independent but they can be optimised to become a good looking graphic that conveys the maximum information. This book skews toward the design end of the range. I don't imply that this is unacceptable, they stay well within the range of acceptable scientific practice, but the bias is definitely present; from the prismatic cover to the dominance of graphics over short, sharp, almost bulleted, paragraphs of text, design dominates.
The initial section is about basics; how to present data, questions to answer about the purpose of the graphic and it's audience. It's thin on details of technique and some of it's examples are not the best in execution. To whit, the use of rainbow scales for continuous variable is generally not advisable since there is no implicit ordering. There are much better scales to use, many based on cartographic principles (e.g. see Cynthia Brewer (2005) Designing better maps). The next chapters are vignettes and case studies organised by content and design constraints - Form and structure, Process and time, Compare and contrast, Case studies and Interactive graphics. The book also has a very nice visual index showing thumbnails of the designs in each vignette and case study. Some of the examples are outstanding, including Felice Frankel's photographs of two colored waters creating contrasting square drops on a hydrophobic grid - a beautiful image that makes a sound scientific point; the two droplets don't mix.
The design of the book resembles that of a tabbed software manual and in doing so reveals the limitations of todays graphics designers who, in their advertising and commercial art heavy design courses, have little contact with the details of book design. A manual is designed with tabs so that short, relatively independent, entries can be consulted to solve a particular problem. Despite the design, and the attempts of the authors, this is not that kind of book. The case studies and vignettes are enjoyable reads, especially the one by Jane Richardson about redrawing a molecular diagram for a paper she was reviewing - I liked her drawing better than the final author generated one, but do not represent descriptions of how to solve a problem. They provide useful ideas and after reading several you gain a feel for how the authors would solve the problem, but patterns for solving a design problem they are not. The tabs get in the way of actually reading the book continuously. A better solution that would have been regular cut pages with color boxes printed to the edge of the page -- it would have provided quick access to sections by the color on the edge and a whimsical note in the preface could have told readers to write the name of a section on the edge of the book over the relevant color.
While I have focused on the shortcomings of the book, it does have many highlights (I've mentioned a couple of them) and the visual design approach it offers is admittedly quite pleasing to the eye. I would suggest that this is a secondary book for scientists or students to read after they have read Tufte or Cleveland. The classic approach to statistical graphics of emphasising data content over appearance is still the preferred approach in scientific papers, but in grants, where salesmanship plays an unspoken but significant role, this book could have much to offer to someone designing persuasive graphics.