Academic print journals can signal some possible pitfalls and solutions to the upcoming market change. When I first went to college in the early 90’s, it was a big deal to be able to easily request paper journals from other university libraries. Flash forward a decade, and I finished my undergraduate degree and have done a significant amount of research without laying hands on a single paper journal. I read them all online.
Many academic databases have PDF copies of journal articles available to paying customers. The services are paid for either per-article, which is pricey, or as a blanket license from the academic institution. This brings all the advantages of ebooks to college and academic audiences. The academic journal market is a hint of where the ebook market could be in the years to come.
Some journals, however, place a digital “embargo” on the most recent publications for up to a year. This is a transparent attempt to keep university libraries subscribed to the print versions of the journal. Such a system could keep print books around for a longer period of time. It could also lead to people being extremely upset with publishers. No user likes the digital embargo.
The expensive print versions made sense when it was necessary (and expensive) to create and store multiple digital copies. University libraries and researchers could only feasibly subscribe to a limited number of journals for reasons of storage space alone. It was a narrow and deep market model. Now, however, the prices could be lowered and marketed to a broader audience – a broad and shallow market model. By continuing to force the narrow and deep model through the use of electronic embargoes, it is going to limit available choices for consumers – and not in a good way. In the meantime, embargoes subsidize the print version for no reason but the publisher’s fear.
This kind of subsidization is going to continue as long as publishers consider the paper version the primary venue and the digital version is considered a secondary venue. I would not propose switching those priorities – not even for academic journals. In today’s world, the preparation for both venues needs to be streamlined and consolidated.
Instead of preparing two separate versions, publishers must reconcile the steps as much as possible. When you do that, you concentrate on the work first, the medium second. If a consumer wants to pay a premium (because of the fixed distribution costs) for a paper version, that’s great. If not, a slightly cheaper digital version lets more consumers buy more products than they would have before. 
That permits a different way of looking at the fixed costs – a view that can adapt as the market changes, keep prices reasonable, and keep authors fed.
This especially goes for the academic journal example – libraries want to have access, not to necessarily read every article out there.