*Almost everything depends on the editor of the journal and his/her prejudices. He/she determines WHO "peer reviews" the study. So if, say, "Nature," has a pro-GW editor, there is no way that study gets out, no matter how valuable or legitimate its research and findings are.
*Peers have their own axes to grind. I once sent in a paper (eventually published by the editor who overrode the reviewer) that was sent to the very guy I criticized. OF COURSE he found my conclusions "unconvincing." More often than not, even "blind" reviews can be figured out pretty easily: you know who is working in what fields, and from conferences, you know who thinks what.
*There is funding on the line. I don't think this is nearly as powerful a motivator as #1 and #2 above, but it is a motivator. If you're getting your $$ from the government, and a new study will threaten that, you're not likely to find much value in that study. As with lawyers and "expert witnesses," the expert witness NEVER comes to a conclusion that is antithetical to the lawyer's position (the guy who is paying him).
I might add that in 25 years in universities, I've never once seen a "pilot" program that resulted in the administration or department concluding, "You know what? That was a waste of time and we're not going to implement it." If it even GETS to "pilot" status, they will always conclude that "it works," if for no other reason than to justify the pilot funding!
Finally---and I'd like other academics to chime in here---there has been SOME evidence that scholarly studies/scientific studies can say one thing, and that the ABSTRACTS, occasionally written by editors, makes it sound like the study says something else. (Michael Crichton pointed this out in his book, "State of Fear"). Well, the idiot news media NEVER reads the studies, only the abstracts!
You may have heard about the huge blow up at the American Chemical Society journal. The editor was rejecting articles that questioned AGW. The ACS members raised quite a bit of hell.
Oh, and with regard to your comments about abstracts; as you know, the word count for submitted abstracts is restricted and some are fairly low. Even if the author attempts to accurately reflect the article content in the abstract, word-count limitations can result in a less complete understanding of the work.
So, even well written abstracts may not tell the whole story.