As an aside, isn't Alexa sampled from a self-selecting set of people who install a browser toolbar of dubious value? Should its figures really be relied on at all?
I'm not sure there is a great way to really measure this type of data. Granted I don't think Alexa is a good point to argue if something is top 3 or top 5. It at leasts provides some sort of indication of overall measurement error. We can probably say it is at least in the top 10 which is probably meaningful enough for the discussion at hand.