Cuil.com is the new search engine that, as ReadWriteWeb point out, got rather a lot of publicity for its launch. Its publicity seems to be based on the fact that it is run by some ex-Googlers, and that it makes some big claims about the size of the index. However, I seem to be missing the feature that will make it a Google killer.
Whilst wanting to be part of the next generation of search engines, it seems to be playing a rather old fashioned game by going for the simple interface and bragging about the size of their index.
Most search engines gave up index-bragging years ago. Beyond a certain number of pages the size of an index becomes quite meaningless for all but the most obscure of queries. If anything a larger index may hamper the results as more low quality pages will be included. It is best to focus on a quality crawl rather than the biggest possible crawl.
Whilst the public love a simple interface (they are simple creatures), it brings nothing new to the market. Whatever way you try to rank the data, whether PageRank or BrowseRank, there is only so much you can do with a simple keyword search: people will continue to use homographs and fail to use appropriate search terms.
Whilst you can only really tell how good or bad a ranking algorithm is by using it regularly, first impressions of Cuil are not good. A simple search for webometrics fails to find any of the three main webometrics blogs, whilst the Statistical Cybermetrics Research Group at the University of Wolverhampton is coupled with a photo of a guy in a turban. No one in the group wheres a turban.
Cuil is definitely no Google killer. These days there are a million and one reasons to go to the Google site besides search, and any new entrant into the market needs to offer something outstanding to break the monopoly. Cuil has nothing.