I'm a pretty regular reader of Kevin Carson's writings over at the Mutualist Blog and at C4SS, the Center for a Stateless Society (I only read his contributions and cannot recommend any of their other writers to date). As a self-described "libertarian leftist" he contributes a lot to a vision of a distributed and decentralized society, one which is potentially more robust economically and socially. His work is a generally well reasoned synthesis of wide range of philosophers and economists; I recommend not only his blog posts but the longer works published as PDFs. In a nutshell: broadly available tools of production and raw materials should make it possible for more people to participate in the economy as producers, to contribute improvements to processes and designs, and to realize the full value of our efforts. Decentralization reduces the concentration of wealth and power and make the system more resilient because the failure or destruction of any one contributor (individual, group, firm) does not have a huge effect on the ecosystem. In such a system, no one producer is too big to fail, contributors leverage incremental increases in knowledge, and everyone realizes the bulk of the value they add to the economy.
Sometimes Mr. Carson gives advice on how get to such an ecosystem. In Steal this Textbook, Mr. Carson identifies textbook producers (especially publishers of college texts) as market manipulators, and suggests a course of action to reduce or remove their influence in the market. He points out that professors both write and recommend texts for classes, and suggests that new editions are produced and printed for what are typically minor changes (which disrupts the market for used/re-sold books). These points are debatable; but let's assume arguendo they're true. He suggests the market power of the textbook publishers can be broken by a coördinated effort to scan and electronically distribute copies of their books; however, I think this is not a reasonable course of action, and the effects may run counter to what he might intend. It's also a recommendation to break the law, which I can't condone anyway.
My second objection is that his suggestion is counter-productive. A wide scale disregard for copyright would undermine (for example) the Free Software ecology; the enforced sharing inherent in "copy-left" licenses such as the GNU General Public License requires the respect of the rights of the authors and copyright holders in the relevant software. Without such protection, such gifts to the community could be appropriated by publishers who would benefit from the value without contributing anything to the community in return. In another example, the works at C4SS (and on this blog) are published to the readers under a Creative Commons "By Attribution" license, which means you can use the work however you like as long as you credit the author(s). The academic ecosystem relies on reputation, and attribution is a crucial component of that calculation. A disregard for copyrights would undermine these and related sharing based environments.
Another objection arises from economics. When textbook publishers lose revenue because of illegal copying, they can to some (large?) extent recoup the loss by increasing prices on legitimate textbook purchasers because there are as yet no alternatives to their products. An example comes from computer operating systems: it isn't massive copying or "piracy" alone which makes Microsoft worried about their hegemony, it's the presence of alternatives to MS-Windows. As long as there were no viable alternatives to using MS-Windows (remember the 1990s), Microsoft could take advantage of their monopoly position to raise prices to maintain their profit margins, and their corporate and legitimate end users had no choice but to pay the rents. Now that Apple is increasing its market share, Linux is taking over data centers and desktops, and distributors like Lenovo and Dell are pre-installing Ubuntu on computers, Microsoft faces an upper bound on what customers are willing to pay for their software, which makes piracy a much more potent threat to their revenues. I suggest to Mr. Carson that it's far more effective for people to support and contribute to open education resources such as Connexions than to spend time scanning and distributing copyrighted texts on the internet. Once there is real competition in the textbook space, the publishers will start worrying about their rotting corpses [being displayed] on [our] battlements. Until then, the publishers can figure out alternate ways to extract revenue from students: mandatory textbook fees per student from universities? Textbook rentals instead of sales (like K-12 schools)? Higher textbook prices? etc.
As I and other comment writers suggest on the article, readers should look for open source textbooks, wiki books, and open education efforts such as Connexions to participate in. Having valuable peer-produced texts to use in education will start the process of forcing the textbook publishers to change their profit model to survive. I don't think it's worth the effort to scan and distribute textbooks; all that will do is convince textbook publishers to adopt the RIAA approach. Far better to have publishers react as IBM did to free software: figure out how to make money by providing added value, either with improvements to the products, or in associated services.