This is kind of chicken-and-egg problem. Today most established journals with high impact factors (= prestigious) are non-open-access. Publishing in or being a reviewer of a high impact factor non-open-access journal looks much better in a resume of a scientist than publishing in non-prestigious open-access journal.
For an open-access journal to become prestigious it needs high quality contributions. But at the same time an author of some important discovery will more likely to publish it in some high prestigious non-open-access journal. Of course for tenured faculty members it is not such a big problem because they have a secured position. But for their apprentices not having high impact publications may become a problem in their future career. And while tenured adviser may force a postdoc to publish in an open-access journal most of them will not likely do this because they understand that this puts members of his/her lab in a bad position compared to competing scientists.
I think that well known scientists should make the first move here and to start publishing in open-access journals. Their work has a lot of traction and will not suffer from being published in some not-so-well-known open-access journal. On the other hand this will help open-access journals to start building reputation.
In some communities open access journals are the most prestigious journals. In machine learning JMLR (http://jmlr.csail.mit.edu/) is (I think) widely considered the top journal. It was formed when a large group of top machine learning researchers grew fed up with the high cost of paid publication and together resigned from the board of a competing publication (http://jmlr.csail.mit.edu/statement.html). All of the top machine learning conferences (NIPS, ICML, UAI, AIStats, COLT) publish their proceedings online too. I think it's only a matter of time before all other communities follow suit, but some communities may lag behind longer than others. The machine learning community was lucky in that a large group of leading researchers decided to collectively make a clean break and support a new journal--the new journal became the premier journal almost over night so there was no slow transition phase.
Don't forget there are plenty of open-access journals with very healthy impact factors (for those who have to be ruled by that exceptionally stupid measure). For example, the top-ranked journal in the Biology category in the 2009 Journal Citation Reports was PLoS Biology, with 12.916.
This is the exception, rather than the rule. Also, remember that the question isn't "is there a open-access journal with a healthy enough impact factor/prestige?" (as if there was a threshold), it's "will I have take a noticeable hit in impact factor/prestige to publish open access?" It's hard to ask young researchers to do that latter.
A wonderful trend people could encourage by making the non-open journals pay for reviews. It'd help the open journals compete.
In there end there shouldn't even be journals, just articles you're pointed to by smart people. Maybe someone would start a ThesisHub for collaboration...
BTW everyone, please release papers in a more useful format than PS/PDF. HTML/CSS is intended to handle device independence and different user needs and would be a better choice. PDFs, which I always read on the computer anyways, are more of a burden than a benefit. In a browser I can run Aardvark or other live editor and make a page readable. With PDF I'm stuck dragging a too-small window over a perfect representation of a useless printed page.
That would certainly work for relatively small or close-knit groups where it is easier to collectively decide to move. I would expect that impact factor has a significant amount of gravity associated with it so even if the leaders in a field decided to move to open-access journals unless the (up and coming) majority also made the same move then the impact-factors of both journals may not change much.
Most people in the EE and CS areas I have seen do. The reason is that they don't invite just anyone to do the reviewing, you have to be deemed knowledgeable in your field, so it signifies that.
You can put it on your CV, although it is not important. Being a reviewer counts as "service" for your annual performance reviews, which is slightly more important. But most importantly, nobody wants to waste time reviewing mediocre papers. Everyone wants to review the best papers, from which they can learn the most while still fulfilling their reviewer duties.
Yes, unfortunately, it is almost more important to reject the bad papers. The good papers will be published regardless of the reviewer. It takes more work when I get a bad paper, especially when it comes from someone famous, because I have to make a convincing case for rejection. I've even gotten the same paper multiple times from different journals (after rejection an resubmission). It's hard work to prevent flawed science from being published.
For an open-access journal to become prestigious it needs high quality contributions. But at the same time an author of some important discovery will more likely to publish it in some high prestigious non-open-access journal. Of course for tenured faculty members it is not such a big problem because they have a secured position. But for their apprentices not having high impact publications may become a problem in their future career. And while tenured adviser may force a postdoc to publish in an open-access journal most of them will not likely do this because they understand that this puts members of his/her lab in a bad position compared to competing scientists.
I think that well known scientists should make the first move here and to start publishing in open-access journals. Their work has a lot of traction and will not suffer from being published in some not-so-well-known open-access journal. On the other hand this will help open-access journals to start building reputation.