The author, Brad Cox, may need an introduction. He is the original creator of Objective-C and therefore indirectly responsible for products like NextStep. Although he has probably been less visible than other contributors like Stroustrup, Coad and Rumbaugh, he has been an influential part of the development of the Object-Oriented paradigm. That alone should give his opinions some weight. However, you need not go on authority alone. While this is obviously not as practically applicable as most computing books, Brad Cox provides some thought provoking insights.
Obviously a book like this cannot be summed up in a few sentences, but let me at least try to paraphrase his central argument. Cox contends that, for the software industry to achieve its own industrial revolution, we need not new technology, but a new economic framework that rewards creation of components in proportion to their use. In a way, he suggests, software production must become more like hardware production - those readers familiar with his earlier work will recognise the Software-IC concept.
In my opinion, the fatal flaw that Brad Cox and others make in their arguments is not that they compare software development to the manufacturing industry, but that they compare it to the wrong part of manufacturing. Software development is more akin to the product design and development phase of industrial production than to the manufacturing. So, while chips are indeed getting ever cheaper and more powerful at an astonishing rate, their development also consumes increasing amounts of time and manpower.
The book is marginally dated (the Internet has moved on considerably since the book was written), and like so many others suffers from insufficient copy editing. However, don't let this deter you; if you have time for a more far ranging book in between keeping up to date with all the latest releases, this one could be a strong contender.