Fascinating - according to that link, the Wolfinger quote stated the opposite, that the plural of anecdote _is_ data. I guess I could see it going either way depending on what you were arguing, which kind of makes the statement lose its punch.
That has been my point several times. Anecdotes are data, they just aren't very good quality data, but sometimes they are all that is really available. Their biggest problem are lack of controls, but then many statistical analyses lack controls; and lack of consistent baselines and alternatives between anecdotes.
That's brilliant. Can I steal that from you?