That sounds about right. It's also what one would expect of a bell curve, which one might see with a sufficiently large sampling of research papers. (If some papers fail on their own merits, perhaps the standard bell curve with 10% failing grades is too generous?)
The L.A. Times is reporting that "about 30% of papers are plagiarized, either totally or in part." As far as I can tell (the writing is poor, but hey, it's the LA Times), they are claiming to have gotten that figure from Turnitin.com. I'm calling BS on that.
Regular readers of this site will know that 1.) I hate plagiarism, and 2.) I think it is far too common. But there is no way that 30% of papers are plagiarized. I would guess that about 10% of the papers I see contain some form of plagiarism.
THE INCIDENCE OF PLAGIARISM. It's not as widespread as the drive-by media would have you believe, according to Unlocked Wordhoard.