Don Park wrote an interesting piece on how competition in the search space might build walls around certain types of searchable content. But that it might also lead to access to the "deep web" and/or richer, easier search of metadata:
How this phase unfolds doesn't have to be as ugly as I described above. More palatable scenario is to open up previously unsearchable areas of the Web and make it available only through a particular search engine. Another approach is to enrich search result using metadata not available to crawlers. Don Park: Search: Great Walls and Silk-roads
Which goes nicely with this Yahoo! Search blog post, talking about a look into library catalogs with the OCLC. Will Google search/index this as well? (the Yahoo! logos slapped all over the place make me suspect no)
But there's more data locked up in the deep web. I'd love to have more of it available, whatever makes that possible. Reviews, listings, recipes, real estate....sheesh, sounds like I'm channeling Marc Canter!
In any case, any sort of walls will I think give rise to the search of meta-engines. If competing search providers have access to separate silos of information -- I would want access to both.