Um, doesn't peer review also include correcting the information?
No. Peer review journals do not correct information. The peer reviewers haven't run the studies themselves, so they don't know what the real numbers are. What they do is give suggestions to the study authors regarding a better or more acceptable methodology, statistical analysis, or what have you. Peer review journals address the acceptability of a study, not the validity.
What span do those papers cover?
Depends on the study and the journal, I guess. Some have been around for many years. They may cover studies from 60 years ago, or longer. Some may be relatively new, and you may only have 15 - 20 years.
As far as studies, they also vary. You could have a study that took a few weeks to run, or a longer study that took several years.
One of the benefits of journals is that they are able to publish studies faster than, say a book publisher. Still, there isn't as much consistency on the quality of studies in some of the longer lived journals, especially in psychology journals. I've seen published studies in peer reviewed journals dating back from the 60s that where terrible. One of the ones that stands out was a one page study. If I had turned in something like that to one of my professor in graduate school, I would probably have been kicked out of the program. Still, it was published.
How many articles are published?
That depends on the journal, and how often they publish. Some only publish once a year. Others publish quarterly.
what proportion does that represent?
Not really sure, but the real problem is how many get by and are never discovered? The percentage of bad studies may be higher or lower, depending on a lot of things. The problem is that we have been sold this idea that peer reviewed is some sacred process that gives credibility and validity to a published study. The truth is that peer review only addresses acceptability, nothing more. In psychology, this more the case. You publish these studies in a peer reviewed journal, and everyone feels that the study is valid. It's not. It's only acceptable by the standards of a small group of people in the same field with similar interests.