This is a key point: That new modes of scholarly distribution will gain a foothold only when senior faculty, those with well-established reputations, start contributing to open-access journal and book publishers.
It does seem now that it shouldn’t be that difficult for a journal to allow visitors to its web site to mix and match articles into customized volumes that could be printed-and-delivered on demand.
As a university press publisher whose titles are available for free on line and in print as print-on-demand titles, I find this argument all but unassailable. From the publisher’s standpoint, the exposure via the web to book browsers is gained with none of the traditional up-front costs (printing and distributing books to stores all over the country). So not only do you reach more potential buyers this way, you do so at virtually no cost. Too few of these experiments have been launched, and have been in effect for too short a time, to demonstrate whether they pencil out, but it seems intuitively obvious that they are more likely to than the traditional print-distribution model.