It depends on what you want to know.
Looking back at the OP, demo boards have
- a score estimator which I believe is a katago score estimator, so quite reliable.
- you can fork the game from a demo board.
Pauses doesn’t really make sense since there isn’t a clock so I don’t know if somehow a demo board had a clock back then.
That said me saying “the katago estimate is reliable” assumes that one is katago, and isn’t leaving weaknesses in a position which the bot can spot but players of different levels might not, namely
It’s not quite the same issue with marking “dame” weirdly but probably similar idea.