What it's trying to do is clear - and the plagiarism example a bit further down is (mostly) spot-on. (It's not unreasonable to mention the Chicago Manual for context there - it's the explicit conclusion-drawing that is a problem)
I don't agree.
http://en.wikipedia.org/wiki/Wikipedia_talk:Attribution/FAQ#Plagiarism_examp... http://en.wikipedia.org/wiki/Wikipedia_talk:Attribution/Archive_9#Synthesis_...
Very brief summary: I think this is a reliable sources issue (the Chicago Manual of Style isn't a reliable source on whether any particular person has committed plagiarism), not an original research issue (it's not original research because the conclusion "by the Chicago Manual's definition, Smith is a plagiarist" is a straightforward logical deduction.)
It's as if the Chicago Manual of Style had defined a plagiarist as an apple, and then someone wrote "according to the Chicago Manual of Style, a Yellow Delicious is a plagiarist". Unreliable source (it obviously contains bad definitions)--certainly. Original research (determining that something satisfies the definition)--not at all.
Also, the example is based on a real case and seems to misrepresent what went on in that real case.