I don't have pre-results to show, but here are my post 02HDs. Maybe just run the bench again - could have had some traffic on the bus or memory tied up elsewhere?
The really good thing to note below is the -1% CPU utilization. So if I have 100 instances of HD Tune running I can effectively turn my quad core into an oct core. Awesome! Best part is, this comes with no additional energy cost - I'm going to apply for some Obama green tech grants to pursue putting HD Tune on everything!