I'll be interested to read the results of the same study this time next year, assuming a similar study takes place. I suspect we'll see user satisfaction levels climb quite significantly. I think this because I believe Siri learns as it deals with queries -- not just about the nuances of different accents and the like, but also in terms of improving its search results. In other words, I think next year's Siri will be far more accurate. This should make it more useful, I think.