In the sciences, most of the primary sources are closed, and so linking wouldn't be too helpful. The material also tends to be so technical that only experts in the field can read it.
As for that class of secondary sources (e.g. review articles and editorials) that is provided for free in a dumbed-down form, then, yes, journalists should provide links.
A custom of providing links would be very helpful in the "chat about science" writing industry, so much of which is based on opinion and political view.
But would it really be that difficult to give me the citation so I can look up the paper in my university's database, or pay for it online if I'm that interested? A proper citation instead of "scientists from the university of blah" is still an improvement.
Even if they just put in the author, venue & volume/issue or, for annual venues, year, that will generally be enough information to find it without a full citation.
Absolutely. For someone interested enough to read a 10-page scientific article to verify a reported fact, $30 fee is usually not a big deal (plus the 5-10% of the population who are in the universities can get it for free). I am surprised the science journals themselves are not pushing for this.
I think it's better to say that in science, many of the primary sources are closed. There's been a huge boom in open access publishing in the last 10ish years, so now well over half the journals I read are open access, and PNAS has an open access option.
Especially when one of the egregious distortions mentioned was from PLoS One, there's just no excuse not to go back, glance at the article, and maybe any linked News and Views type articles.
As for that class of secondary sources (e.g. review articles and editorials) that is provided for free in a dumbed-down form, then, yes, journalists should provide links.
A custom of providing links would be very helpful in the "chat about science" writing industry, so much of which is based on opinion and political view.