Destroy service should invoke unit's stop hook, verify/investigate this is true

A report came in on irc, that destroy-service wasn't killing off the underlying service processes. The report also indicated the unit agent was surviving, this needs some investigation. A preliminary look at the code suggested that we break service-relations, and destroy the service, triggering the machine to destroy the service deployment (which should kill the unit agent process), but this won't in turn trigger execution of the 'stop' hook.