Doesn't seem to matter whether the author already knew that she had a knack for talking to strangers without appearing crazy, or held with the "general assumption" and was genuinely surprised when the experiment turned out well. The important thing is that it worked!

Now, perhaps it will not work so well for every person. Perhaps in the general case people will instead recoil in horror at being approached nicely for small interesting favours. Perhaps a more universal lesson to take away from this is that success is more likely if you play to your strengths, and push every advantage at your disposal. The author is a Cambridge Law graduate, so one might safely presume reasonably skilled at speaking - I'm not exactly sure how surprised she was that she could engage strangers in professional conversation at a coffee shop - but the experiment is available for others to replicate! I'm not sure what parameters would have to be changed for the interaction to take on a "dramatically different character" though, maybe don't get drunk beforehand.