Not only does this discussion apply to printed books, but also to the music industry (one main difference being the music industry generates boatloads from live events, which the book business fails to do).
So, if one believes in Schumpeter, then the business of publishing/producing must be broken down. At then end of the day, there is still plenty of money to be made. It just needs to be made slightly differently.
In an attempt to be overly simplistic, a successful business should be singularly judged not on revenues or market share but on cash flow. So, at a basic level, its okay if revenues go down by 50% so long as gross margins go up by 2x (ie. the same gross profit) or if operating expenses go down by a substantial factor (due to a simpler business model), or some combination of the two.
So, the question is that if only 1 of the 2 jobs that the publishers do is getting disintermediated, why is it the end of the world for publishers? Yes, things are changing. And, yes, jobs will be lost (and gained). But, with such a critical asset (the editing expertise), these companies should be far from doomed if well managed.
Pulled from a recent New Yorker article on the Amazon Kindle is this tidbit.
The newspaper industry, [Russ] Wilcox [founder of E Ink] figured, was a hundred-and-eighty-billion-dollar-a-year business, and book publishing was an additional eighty billion. Half of that was papermaking, ink mixing, printing, transport, inventory, and the warehousing of physical goods. "So you can save a hundred and thirty billion dollars a year if you move the information digitally," he told me.And, according to Wikipedia, the book publishing industry employs 25,000 people in New York.
A digital book on Amazon still sells for nearly the same price as a printed paperback (or roughly a 15% discount). Based on that and the above estimate of costs in the business, it would seem that gross margins on that product have skyrocketed with the advent of digital delivery. Higher gross margins means more profits (one would think), especially if operating expenses remain flat, which we might be able to assume.
I'm pretty sure that the vast majority of those 25,000 people are not actually "printing pages" or "making paper" or "driving trucks" for the delivery of books, but rather they are likely engaged in a variety of value-added publishing services such as editing, discovery, marketing, etc, etc. Those value-added services are still needed by authors regardless of how a book is distributed.
So, rather than worrying about the publishing industry being destroyed, we should think more about the printing industry seeing declines. Wait, hasn't that already been happening for at least a decade?