While we have not enabled experiments for users with DNT set to '1', we
were sending snowplow events and recorded the users within the control
group. This is not only against the idea of DNT, but also influences our
experimentation data.