I know that what I will say is going against the main stream, but alternative points of view are important as well. As we can all see by the incredible number of new Open Access journals flourishing daily and harassing scientists and other scholars via filling email boxes with invitation to submit a paper there, Open Access it now very fashionable. And as we can read in Nature, there is a global movement even in funding agencies to promote Open Access, as a wider way to spread scientific results.
Of course, nothing can be said against the fact that, once published, we all would want our results to be accessible as largely as possible.
However, there are some risks as well.
With the former publishing model, in which Authors submit an article to a peer-reviewed journal, the Authors who do the experiments and write the paper do not have to pay more. If accepted, the paper gets published in the journal, and it is readers who will pay via (individual or institutional) subscription, or with a pay-per-view system. I don't want to discuss here the question of the pricing of the subscriptions, which can reach extremely high levels for universities which often have to maintain active tremendously large collections. Money is sparse, so often universities unfortunately have to make choice. But in this "classical" model, the interest of the journal is to only accept the best papers, so the reputation of the journal will stay high, and consumers will be willing to pay to keep their subscription.
With the Open Access model, it is a bit different. Authors submit an article to peer-reviewed journal. But, if the paper is accepted, the Authors have to PAY to make sure that the paper will be in Open Access, meaning freely accessible by anyone anytime. Great, but nothing is free, it is in fact the Authors who pay for that. Meaning, not only the Authors have to do the experiments, to write the paper, and to finally pay to have it published. Given that Reviewers work for free, the money goes all in the journal. I am not saying that the journal does not have fees and people to pay. Nor I am saying that it is a bad system. There are some excellent peer-reviewed Open Access journals, the first names coming in mind being PLoS ONE (which initiated the movement) and the Frontiers series. But still, in such a model, the interest of the journal is not necessary anymore to insure the highest level of publication, but to get the most papers published, since the money is not coming from the copies of the journal purchased (linked to the intrinsic quality of the papers inside the journal and to the overall reputation of the journal), but from the number of papers published. And hence, here is the danger to go for quantity over quality.
Having said that, I myself publish a lot of papers in Open Access journals, so I am not fundamentally against this idea (not at all). Just, I believe we still need to find the optimal way to articulate the Open Access model of publishing with the more traditional systems, in order to keep the quality of published science: only keeping the highest quality for publishing research (whatever the field, whatever the Impact Factor) can pave the way to human progress.