Ironically, this is due to the lack of a text content industry service that crawls for copyrighted works -- like the ??AA.
They're having such a hard time because it's not automated yet and they care too much. Providing for free (or with attribution) is fine. Plagiarism is not.
In other words, RIAA shouldn't employ MediaSentry; MediaSentry should sign up content producers for cheap p2p checking and a MediaSentry analog for text should sign up content producers for cheap copy checking.
New model: bloggers and newspapers make money from ads, spend money on content checking. If your content can't support its own uniqueness via ads->anti-plagiarism, you lose.
Thoughts?