I ran two at a time for four weeks each with a bit of patience, so this is not theoretical. On the numbers, whether verification exists at all explains most of the complaints here. The trend was steady fairly clearly long after the novelty had worn off.