Unneeded check if rule returns false

[BehindTheMath]

When a rule is checking any additional rules added with .and(), it checks if one of those additional rules returned false. However, it appears that the only possible return values for a rule are true or a string, so this check is redundant.

[rtheunissen]

This allowed for rules to simply return false, but in hindsight I think we should enforce a string return. I have a feeling that there was a good reason for supporting a false return but there is no record of that. Notice the failing test that was also removed in efcb2cc