I'd love to see some new search engines pop up and break up Google's monopoly. I'd like to see Apple release a better designed search engine, someone make an open-source search engine, IBM make a search engine with the tech behind Watson, etc. Right now there just aren't many good choices. Bing is basically Google with a face-lift.
I use Blekko. I like it, but it needs a lot more people creating topical lists of sites before it really offers something unique.
That already exists in a sense. You could start with a bunch of ASF projects... Nutch, Solr/Lucene, ManifoldCF, Droids, Hadoop, OpenNLP, Tika, Mahout, UIMA, etc. and build a reasonably good search engine. The problem isn't writing search code; it's scaling the darn thing up to "Internet Scale" and other things that get ya. Can you imagine how much hardware and how much bandwidth it takes to continually crawl the web, download, parse and index pages on the scale of a Google?
The other problems are things like preventing spammers from gaming the system, etc. Whether or not a search engine where all the algorithms were public would be easier to "game" is, I suppose, an open question. I think most of us intuitively feel that it would be, but maybe not.
No it would not be reasonably good compared to Google. It wouldn't even match what AltaVista used to be (given the relevance algorithms used in these products).
Note that I said start with those projects, and not end with them. Of course it would take more work to get in the same category of relevance as Google. But the point was that a significant portion of the code needed to build an "open source search engine" exists. But, even if it all existed, the problem would still be hardware and bandwidth. You can't easily build a Google-like without some serious financial backing, even if everything is open source.
I use Blekko. I like it, but it needs a lot more people creating topical lists of sites before it really offers something unique.