Willy,
I'm not sure if it's merely a matter of syntax (although I imagine that could play a part). Subject matters, topics, and interests would have to come into play, sooner or later. My own suspicion is that nobody could be given a fair chance to pass (or fail) the test on the basis of just one or two conversations. The sense we get of who a person is, over time, would provide the best clue, I think. Of course, even then we could be mistaken.
Violet,
I don't think that a man persuading someone he's female (assuming, again, a text-only environment) would mean that zie is female, no. I think it could, however, mean (amongst other things) that he's good at persuading others that he is, indeed, a woman or that he possesses character traits traditionally associated with feminine psychology (my next question: is there even such a thing as "feminine psychology"?--this is what, to some extent, a Transgender Turing test would seek to answer).
Tamara,
Heh. I see your point, about that particular double-standard. I've been known to say things like, "Yeah, pressure: third down and two strikes already." But, I assure you, I am not a woman (and I know women who would laugh me out of town for making that kind of statement). Funny how such things go.
Thanks for the replies, all.
Love,
CJ
