Many students (and some researchers) do not care about the name of the journal that houses the article. They care about the article itself. This viewpoint will continue to grow as they get older. More and more patrons will find articles through Google Scholar and other databases instead of browsing the current issues of Science, Nature, JACS or whatever.
Believe or not, my library is going through a cancellation discussion for the first time in 18 years. Our faculty and students have not had to worry about cancellations in quite some time. Thus, they have been shielded from the growing STM serials crisis. The faculty and the students are not the ones who pay the bill for the information. The library pays (well, the University actually) and the patrons are the ones who enjoy the benefits of the access to all of the subscriptions. This strange economic model has been well researched by Mark J. McCabe.
I wish I had time to read all of the 87+ books - and - reports that John Dupuis posted to his blog. I've already read a bunch, but there is so much great insight to be gleaned from all of these. (I wish I had more time to read...)
I imagine a future world where scholars post their work on websites/blogs -- it gets critiqued by a variety of scholars, and the articles get rated. The scholarly articles that get the highest ratings, citations and links get brought to the top of search results. As it stands now, articles do not HAVE to live within the imprimatur of a journal brand to have an impact. They can live on a website, in an IR, at a preprint server, etc. In the future, more and more and more scholarly articles will simply live on the net as stand-alone items without being housed in a journal. This is where the scholarly article market is headed.