Elsevier charges authors to make items "Open Access", then charges to read them.
blogs.ch.cam.ac.uk
blogs.ch.cam.ac.uk
Edit: Yes, it appears to be true.
http://www.sciencedirect.com/science/article/pii/S0168900210...
When I first clicked the link: (a) no open-access logo, (b) button offering me the article for $35.95.
I did not at the time know what the open-access logo was supposed to look like, so I went back to the article, found it, and clicked the link again: (a) open-access logo, (b) article in its glory.
My guess is that in response to this article someone fixed something and the article is now properly open access.
This is the link that is (apparently incorrectly) used as an example of an OA article in the email that seems to have led to most of the confusion.
Unless I misunderstood the blog post, when the author contacted Elsevier, the company's Director of Universal Access wrote back on an email that 'open access' articles can be distinguished by (quoting from the email) "an open access symbol and link in the top right of the article. For an example see here (http://www.sciencedirect.com/science/article/pii/S0168900207...)."
As I write this, that article costs $35.95 -- that's pretty convincing evidence, no?
Edit: as some comments below posit, it's also possible that the Elsevier executive provided the wrong link.
It seems possible that the response just had the wrong link, and they article they are charging for isn't open access. http://www.sciencedirect.com/science/article/pii/S0168900210... is clearly Open Access, has a clear mark of that, and has full text available for free.