Why don't new search engines build off of something like http://commoncrawl.org/ instead of Bing's API? Is it purely cost/resources or a combination of freshness and steeper learning curve?