Repeating what I just posted on Google+:
If this article http://www.sciencedirect.com/science/article/pii/S1751157716301523 is even close to being right–and I have no reason to doubt it–there’s finally a good figure for how much we undercount OA by omitting hybrid: 2.8%–for 2013, 13,994 hybrid as compared to 493,475 gold (in DOAJ journals, that is, not including 188,000-odd in “gray” OA journals: include those and hybrid’s just barely over 2%: 13,994 compared to more than 682,000).
So: next time somebody says “but you’re not counting hybrid articles,” a reasonable rejoinder might be “So?”–and, of course, a link to this [fully available] article.
And a tip of the hat to Valerie Hawkins for pointing out the article…
You’re welcome, Walt! Open access is the way forward!