> - What they found is totally expected and isn't something I'd worry about: it's kind of expected that if one guy warms up with 10 questions done himself, and the other guy warms up with 10 questions delegated to AI, then the second guy is going to do worse on the final 10 questions (because he didn't warm up).

I agree with your general conclusion, but with this point, the solve rate for the AI group once AI is removed is lower than the initial solve rate for the other group in all the experiments (or at least definitely in 2 of 3), so I don't think it's explained by a warm-up.