I don't agree with your interpretation of the article. It seems clear to me that an outsider that doesn't share the community beliefs would be considered most likely to break with community norms and steal money, even if this belief is wrong. It has nothing to do with being considered 'worse' than a rapist. Even though an outsider might also be considered the more likely rapist ... these biased probability assignments are completely different than believing an innocent-so-far atheist is worse than a convicted rapist.