Yes, it's a rather strange objection to make... How could Google possibly use data on its employees to disprove that brainteasers work - if they weren't using brainteasers at some point?