-historical data provides huge scale of information well beyond what you can get from the current data set. This is true for figuring out searcher intent, spam patterns, etc. It’s quite possible Bing won’t benefit from the scaled historical data because Yahoo search is on entirely different infrastructure and there may be no way to import it into their existing historical data storage mechanisms.

-Current searcher data provides lots of things at scale — an indication of current trends and interests (a sudden spike in a topic can feed into what should trigger news results in web search, for instance), searcher intent, etc. It would absolutely benefit Microsoft to have 4-5 times the search volume for this.

-Experimentation. Varian says Microsoft could do experiments just as easily as Google does, but a 1% experiment at Google is going to provide substantially more data than a 2% experiment at Microsoft’s current volume levels.

You can see the difference in data value based on searcher volumes with the keyword research tools the engines provide Compare the data you get from Google’s adwords keyword tool and Microsoft’s adCenter Excel plugin. In many cases’s, Microsoft’s data is too small to be useful.