The problem with 6 is that often the PDF of a research article is behind the paywall of the journal that published it. I agree that news organizations often do a horrible job of citing their sources for science coverage though; they could at least link to the journal's webpage for the abstract.