I would like to point out that the inner complexities of a search engine are more biased then anyone could possibly know. In addition to word counts, word placement, search hits based on previous searches for the same information, Google has been known to add their own metadata to certain web sites (keywords that are not on the page itself but attached to their copy of the page) that are used to find sites that may not list their purpose in text, or use commonly used text to refer to themselves. This is done by your favorite infallible device - a human being (working for Google, that is).

There was a paper referred to by a professor of mine (though he admitted that it might be urban legend) about a group of researchers who set up several automated, focused searches and ran them through Google and were able to decipher Google's search strategy. Then, the lawyers got involved. Too bad.

One should look into certain search sites which use user feedback to determine the relevance of a search term. In that case, a more egalitarian model is possible. It's interesting stuff.