I read some posts in the past about this subject as well. Jason Kottke has written a nice weblog post covering some thoughts on the Google browser and other things he expects Google to do. I would love it if Google was going to invest in the Mozilla project. All we have to do is keep an eye on the official Google weblog I guess.
(Of course, when I think of Google joining the Mozilla project I'm thinking about more CSS support and a better well-formedness parser, not really about Google interfaces.)
Posting comments isn't possible with the staticize plugin, apparently. Probably due the fact Mark hacked mine quite a bit (preview, validation). Mark! ;-)
The people from Google have absolute the knowhow about the structure of webpages. See the guidelines for webmasters.
Sometimes I do a search on my own website to get a better structure. The webmaster is like an architect; You can build a beautiful house, but the bathroom is where he is for practical reasons. So, if we search something on the web, the structure of websites must be clear.
And maybe can Google break the IE-monopolie.
But then again... isn't Google an upcoming monopolist? I don't really bother at the moment, but I must confess that I'm a little amazed that I haven't heard of any open source Google-alternatives.
Robbert, how about mozDex? Ok, the search is down at the moment, but the search was actually pretty good when it was up.
Oh damn, forget the stupid post above, I'm so ashamed. It's not an open-source engine, but an open search engine for a fair index.
Brainstorm. What if Google turned the Googlebot into a browser for people to use? It could then spider pages people actually go to instead of pages that barely get any hits (even if they're linked to a lot). Something like this could be useful for search engine rankings.
There could be something in the browser's privacy preferences where a person could configure how this works. Perhaps they could simply turn it on or turn it off, or just turn it on for certain sites. The browser could also be made (for privacy concerns) to not spider a page (no matter what the users settings) if it's behind HTTP Authentication or otherwise protected page. There's a lot of ways privacy concerns would be removed, and still have this working.
I'm probably only touching the tip of the iceberg of ideas for this concept. I bet Google could become a force in the browser universe with something like this, especially since everyone's knows Googles' name.