OK, but then Google already has that problem now. If they receive a DMCA notice against your pre-registered content they must act. But I'm sure their algorithms would prove that pre-registered content that is not available elsewhere to be extremely reliable. Keep in mind that Google's index probably has basically everything that is out there.
I believe a downside to my proposal is that, if it went into effect, honest publishers would in essence be forced to pre-register, just to be safe. I can imagine some would complain about that.
The DMCA has issues but at least disputes end up in court, how do you think an ownership dispute would play out if private companies were in charge of making decisions.
With regard to the notion of a private company controlling copyright issues, I'm a first believer that Google, big as it is, is not the Internet. An honest effort to clean up their index should not be confused with an attempt to censor anything. Taking the flip side: prove (via a simple API) that your content is indeed original, and then you'll be welcome in our index. Needless to say, only a dominant industry leader could take such a bold attitude.
Why doesn't Google allow, indeed encourage, a pre-publication process? As I see it, this registration would allow me to "predict" that my 5000-word essay, which can't be found anywhere on the Internet right now, will appear on example.com/article within a day. Almost beyond any reasonable doubt that should bring lots of credibility to my allowed site(s) and accurately destroy the credibility of unscrupulous sites that simply copy and paste. By the way, this could also apply to works placed in the public domain [1], and copyrighted material that will not (or should not) appear freely on the web, as is the case with the OP's book.
A refined API would speed up the process and allow authors to attach copyright restrictions to the resource in question. Needless to say, this process could be extended to other search engines, but it's safe to say right now that if Google alone allowed something like this it could rectify many problems in SEO.
I guess pre-registering heavy data such as images or video may not be reasonable yet, but plain text (even excluding HTML and CSS) could be a great beginning.
Thoughts?
[1] Even if a work is placed in the public domain, it would be reasonable to give the author's original version due prominence in the search results.