The results are in! A few thoughts:
I saw inconsistency as the most defining factor of fakes. I felt that if a post was inconsistent in its internal ideas, it was more likely to be fraudulent. For the same reason, when I sought out to write my own entries (A2 and C2), I strived for internal consistency. The underlying assumption here is that someone who is an authority/believer of a certain ideology is more likely to do the intellectual legwork to make that position consistent. Whether or not this is true is another story entirely.
My strategy for writing as a Christian was emulating the positions I had before I left the fold. I really felt those positions when I was a believer, and they were pretty simple to regurgitate. I paused a bit about deferring to others about opinions of morality, but I am a genuine C.S. Lewis fiction fan.
Turing Tests (seem to) make people overly-skeptical. In the comments of various posts, people found really incidental reasons for skepticism (footnotes, undeveloped positions, and questionable allegiances). It is perfectly understandable given the format, but surely those heuristics aren’t useful in finding frauds. I’d be interested if others had criteria that operated more efficiently.
Overall, I think the ITT was a really fun and eye-opening exercise. The anonymity and possibility of fakes made every claim feel like a potential landmine that could backfire and give away your fake (or genuine) beliefs.