I think that the narrator function in the One Night Ultimate Werewolf should have its length customizable. In the app used with the physical board game you are able to increase and decrease the pause for each section of the night phase.
In the TTS version it seems the narrator just has a preset length for each phase and in some cases this length is too short, only about 5 seconds. Which is not enough time for the robber to easily swap cards and check their new card.