I did a test yesterday with my Suunto: two laps in each direction at South Oval. I was half expecting to find nothing because I wasn't running that fast but it measured longer clockwise (520 and 530) compared to anti-clockwise (480 and 500). Not much data but it seems to match what you found.
Spooky and a bit confusing - Matt and I were slower clockwise suggesting under measurement that direction? No idea which direction coriolis would impact. Really hoping Blair will wade in on this sometime soon.
I thought slower would imply it's measured the distance longer, following from pace = distance/time. Read some conjecture elsewhere on the web that it was error from the movement of the arm you carry the watch on.
At this point I'm wondering if I'm a big enough deal to warrant a Blair interjection - and I agree, the way my watch works, it gives me greater distance in the clockwise direction for the same time.