In your parallel code sample, hang2 is only 1000ms instead of 5000ms.

This is very cool though!