I'm curious, because I don't really know this system, if other scientists feel this way. I have some inkling that journals can often publish erroneous papers (maybe I got that idea from Nate Silver?), but that would be so very unfortunate for Science's reputation. I hope it's not that bad.
Journals like to publish sexy/novel papers. That can lead to them publishing some things that are "hot", but not necessarily well proven. The arsenic bacteria from a couple of years back is a good example [1].
In their rush to publish, solid work which may not be as sexy can get pushed to lower tier journals, whereas "novel" things that aren't as well proven can get elevated to top-tier journals. At this level, things aren't necessarily the pure meritocracy that we like to think it is. Then again, these are for-profit publishers. There is nothing that says what they have to accept, and they can choose what they want to publish. The only thing that will change the system is large-scale or big-name boycotts like this one.
