Peter Murray,  Assistant Director, New Service Development for OhioLINK: The Ohio Information and Library Network has an interesting post today analyzing what he describes as "the complex ecosystem of the textbook marketplace" (which he later characterizes as a "mess").  Essentially, he recaps some commonly held perceptions (textbooks are too expensive, the "textbook" model is broken, publishers are profiteering with impunity, students are rebelling, etc.).  He begins his post  with familiar data points from a widely circulated 2005 GAO report that suggested that that the cost of textbooks have risen at twice the rate of inflation.  He goes on the make the point that digital content initiatives have emerged, but are competing (the publishers' collaborative initiative called CourseSmart vs wholesalers' initiatives such at the Universal Digital Textbook).  In the end, he wonders why, like in the music space, technology has not pressured down the cost of textbooks.

"Among the many unanswered questions in this ongoing exploration is why technology has not dramatically reduced the costs of the materials themselves."

IMHO, the answer is simple.  The "materials" we're talking about are high-value--and not easily reproducible at the same quality level as that which textbook publishers produce.  Intellectual content, like that found in peer-reviewed textbooks is still a scarce resource, contrary to public opinion that's quick to discount the value of textbooks.  Plus, it is not just the content itself that has value in a textbook, but also the organization, the pedagogy and ancillaries that combine to create the learning tool.  In courses where textbooks (print or digital) are used as part of a curriculum of study, students tend to learn more, with less effort.  This is not to say this is universally true, but study after study has shown that students report less angst over the cost of a textbook when it is actively used in the course  as part of a larger  educational experience.   Textbook substitutes--including those germinating from technology tools such as Wikipedia--have made little impact in part because the source of the content is unreliable (underscoring the role of the editor in the publishing process), and in part because the content and organization is hard to use when the goal is to master a corpus of content tied to an outcome like a nursing degree, for example.  This is not to say it can't and won't happen, it is just to say that for now, for better or worse, the traditional, peer-reviewed, four-color textbook written by a qualified academic with credentials in the topic at hand is still the best resource going for most academic courses.