I, as one new to the idea of Activity Theory, felt I didn't get as much out of this book as I should have - many issues were glossed over as if they were assumed to already be familiar to the reader. A more thorough grounding in the ideas of Activity Theory would probably have been beneficial to understand the finer details. As far as I got it, a summary of the message of the book is:
People do what they do for a purpose, and one will make a better computer system if one knows what the user's purpose is.
To me that seems neither controversial nor particularly different from what ethnomethodology purports to do. I found the descriptions of actual systems analyses in Part II to be the most interesting and useful material, but they were also so different in execution from one another that I was left to wonder whether there in fact is any *method* to Activity Theory, or if it is simply that idea that users have reasons that then has inspired researchers to make analyses in whatever manner they felt like.
Part III contains a long and rambling philosophy of Activity Theory in the light of the poetry of Mandelstam, extended with psychedelic diagrams. I really felt the book would have been better off without this chapter. Sadly, all the authors' diagrams tend to be more confusing than illustrative and perhaps that also speaks for the less than precise nature of the theory in that there is no well-developed notation for describing one's findings.
The book was produced in the amazingly short time of a year from idea to print, and while that speaks much for the discipline and ability of the authors, I think the publisher should have spent a bit more effort on proofreading and better-quality images.