Energy : Journal costs - what's the answer?. Alternatif Energy

Sorry for the brief break in posting - real life obligations sometimes make it tough to blog as frequently as I would like.

The Nature Publishing Group is going to launch another five journals this year.  University library subscription costs for each of these are going to be around $5K/yr.   Other journal publishers are making similar moves - the ACS has launched three new journals this year, including an open access journal that sounds like it's meant to be a direct competitor to NPG's Scientific Reports.

On the one hand, these journals wouldn't be launched if publishers didn't think they could at least break even, meaning that someone somewhere has done a marketing study suggesting that there is sufficient demand out there both from authors and would-be subscribers.  On the other hand, it's hard for me to believe that the market can really sustain continuous growth in the number of journals, especially when this implies a similar growth in the number of requests to review papers (for free of course) from all of these editorial boards.  

What is the endpoint of this proliferation of journals, especially when many university library budgets simply make it impossible for those schools to pay for institutional subscriptions, and the pool of qualified reviewers is not similarly expanding?  In the long term, it seems like services like the arxiv have to win, perhaps with some kind of post-publication review/commentary.  However, the reward structures in place (i.e., the emphasis on particular "high impact" journal publications in hiring and promotion) put in place a huge barrier to change in that direction.  This is another area where I worry about the inevitability of a greater bifurcation into "have" and "have not" institutions, something that has a certain internal consistency but is probably long-term bad for creativity in research.