Thanks for the thorough comment. I don’t disagree with any of your points. As you way in your last paragraph, I believe that federated search complements Google. There’s clearly a market for both. And, as you also say, I don’t believe that all content publishers will make their content indexable by Google, especially the large volume of non-free content.

Thank you for yet another insightful post.
You mention two reasons why makers of federated searching tools should not be alarmed by Google’s dive into the deep water of the deep Web.
I believe that Google could make great strides towards overcoming your second point, scholarly materials being mixed up with non-peer-reviewed documents of unknown quality, by incorporating its deep Web retrieved records within its Google Scholar service.
This, albeit not providing a 100% solution to this problem, would reduce it significantly and enrich Google Scholar content.
The first point, incomplete coverage of relevant material per content provider or repository could also be negated with time, providing that we see a settlement between scholarly journal vendors and Google along the (unclear, at this point in time) lines of the Google Book Search settlement with publishers.
Once Google is licensed to include full subscription-based scholarly article repositories within its search results, compensating authors and publishers in an agreed upon manner, it will enable Google to offer comprehensive deep Web content with unquestionable authority.
Having said that, I do not believe that these future developments would render current federated searching tools obsolete.
Federated searching tools, as the information world as a whole, would have to evolve and offer services and content which Google can’t or won’t provide.
Human quality control and indexing of select article databases which provide better precision than Google’s deep Web automated tools could be one direction, as well as indexing local content produced by the buying institution, which is sometimes not on the Web at all, let alone possible for indexing by Google.
Plus, there will always be a number of publishers unwilling to have their content indexed within a free search engine which would make the commercial federated searching tool the only tool able to retrieve their content within combined result sets.
In short, I see this as a great step forward for Google, with the potential to lead to gigantic steps in the future, but I also see the commercial federated searching tool as a product which completes Google and will continue to do so in the foreseeable future.