I've been worried about Google for a long time. If they don't index something or mis-index it, it becomes essentially invisible.

I also have an essay on this (actually two linked ones, here's the first):

Google and the Dissemination of Knowledge

I spent much of my professional career developing search and retrieval techniques and Google uses really primitive algorithms. The measures for retrieval are relevance vs recall. High recall means you deliver lots of stuff, but very little of it is useful. That's Google's model. Because the response time is so quick and it is easy to scroll through 10 or 50 hits people find something to look at. Studies have shown that people will take something in the first 20 items returned (this is true for library research as well) regardless of whether it is really what they need.

The poor recall means that you don't know what Google didn't deliver. Since the bulk of the searches are for trivia and shopping people tend not to mind.

Just try to find images or other non-text material to see one of the major failings.

I think the popularity of Wikipedia is because the articles have high relevance, even if they are sometimes inaccurate.

Policies not Politics
---- Daily Landscape