"so called "peer review", to be honest, is a broken system..."
To misquote Churchill:
Peer review is the worst system for scientific publishing... except for all the others.
To misquote Churchill:
Peer review is the worst system for scientific publishing... except for all the others.
In the pre-internet days, peer review like this prevented some bullshit from wasting the relatively scarce resources used by paper / USPS distribution and storage. Today it is just goofy, and only serves as a weird sort of evaluation procedure of scholars who rack up "scholarship points" just like some weird game of monopoly.
What about all the crap that gets "published" electronically, you ask? Almost by definition a "subfield" is a collection of people able to evaluate each other's work; if everybody in a subfield gets emails when a new paper is available, they ALL evaluate it via the internet (chat/ email/ etc), and they are all peers, and we see the old school scholarly distribution system to be parasitic ...
.... except there is no other way to rack up "scholarship points", so people continue to submit to the old school journals with their crusty old reviewers and their bullshit.
As to the purpose of peer review and whether it's still needed: Biology/chemistry is huge and there are probably 100-200 journals that each publish ~50 papers/month that are potentially relevant to my work. I skim them all (and use e-mail alerts to notify me of anything especially relevant) but I certainly don't have time to read them all with sufficient depth to be critical. As such, I rely on editors and reviewers to pre-screen papers and only let the highest quality research into the most-prestigious journals. The system is, of course, flawed in many ways... but, if nothing else, the competitive-ness of the top-journals "keeps the riff-raff out". With the exception of articles that are directly relevant to me, I only take a close look at papers that appear in the top ~5 journals because most of the lower-tier journals are pretty boring. My behavior is pretty typical, from what I've seen in my field.
In a perfect world I'd have enough time to evaluate everything on it's own merit, but I don't (no-one does). But if a paper has made it into a good-to-great journal, that at least means that (a) an editor thought it was interesting and (b) three or four respected scientists read it carefully and thought it was pretty good. As you pointed out, there are lots of places where this system can err, but overall it does a pretty good job and serves my needs well. It's sort of like the point-system on HN: good comments sometimes get missed, and bad comments sometimes get up-voted... but overall it's a pretty effective way to help people separate the interesting from the banal or trivial.
[1] http://dictionary.reference.com/browse/paraphrase
[2] "It has been said that democracy is the worst form of government except all the others that have been tried.”" -- Winston Churchill
PS: Don't rely too much on dictionary.com, it's not the best. E.g. the New Oxford American Dictionary correctly gives the second meaning of "paraphrase: a rewording of something written or spoken by someone else".