That's what I meant. The results show that the DO box is fast enough, but a slow result doesn't indicate that you can't saturate traffic. It's really hard to test without real-world traffic, haven't found a reliable way to do so yet.