I feel the major areas where good visualization adds value is - avenues of knowledge discovery, information presentation or simply context exploration where one does not know beforehand what insights data may reveal or the extent of relationships between data points. It can be debated that good visualizations could possibly increase cognitive resources available to our minds and decrease conceptual load that has to be dealt with at one time. Visual analytics is an invaluable tool for divergent thinking as well. Coming up with good visualizations is hard but its impact can be immediate and convincing. The visual impact is probabaly more because visual pathways have higher bandwidth at the systems level directly to our reptilian brain whereas logical reasoning resides on the neocortex. Sometimes we just have to see to believe.
The book does an excellent job of visiting various applicable domains. There is enough variety to keep you occupied. Its covers visualizing social datasets(link-node), hierarchical datasets(trees), categorical datasets(relational), time-varying datasets and other most commonly found datasets in scientific and information visualizations. Most chapters are well written with plenty of visuals. If you are an ardent fan of visuals you probabaly have seen everything in this book at one place or another. I think a significant part of the information is derived from papers or other well known publications. You can find almost anything by taking the visualcomplexity website as a starting point. However, the book can save you the research and is an excellent introductory text to expand your knowledge on visualizations.
That being said, I feel its little expensive for its size and as most content is there on the web already, O'Reilly should have listed it at slightly lesser cost. In case you are not going to buy, its a 5-star book no doubt.