On a more 'psychological' aspect, I got another example. I don't remember the exact situation, but there is some 'lg uhc' for Minecraft in Twitch where in short it's a werewolve game in real time (or an Among Us game), and I remember TheGuill (a player and streamer) which got information from outside the game (maybe with chat) from Libe (another player). And he shouldn't have access to them.
And yet, TheGuill said in stream in front of more than 1k viewers that he really think it would not be correct for Libe to lie on these informations.
But, both player are clever, just by the game they are playing. But that reasonning doesn't make sense for TheGuill. What TheGuill want, is to have control of what is going on. He doesn't want to use the extra information from outside the game, but he wants to know the 'role' of Libe. And no-consciously, he will use it.
Libe on the opposite, most of the time will just not think that far and what he said outside the game will be true. But why can't he punish TheGuill for using this kind of information as his advantage?
And once TheGuill start to think something fishy is going on, since Libe has a chance to lie, he will begin to doubt about these informations, which is perfect since TheGuill should not have access to them. So lieing has only advantages.
And I think this example translate perfectly to Hanabi, where people don't want to use information in chat, but do it anyway. And where having a good portion of the community lieing on purpose could only be benificial if everyone agree that giving and using information in chat is bad.
For anyone who knows what happen to Libe, it was before that happen, and it's not the subject anyway.