The original running Garmin was a GPS device you strapped on your arm and sent its info to a separate watch that you wore. I had one of these and it was very inconsistent at measuring distance. We bought one of the first watch-type devices, and that's what I've used for my testing along with my hand-held device.
A few years ago I asked people in a road race to report their GPS measurement and their model #. From those reports I didn't see any significant difference between the different models, older and newer. But in that race the view of the sky was mostly unobstructed. I suspect the improvements with the newer watches is their ability to hold onto a signal in less than ideal conditions. However, with my very old watch the only time I've had signal problems is in locations with tall buildings. But maybe with the newer watches the difference is the definition of "tall."
The posted article is actually quite old and has been posted on the bulletin board before. It is a bit surprising to me that someone who has access to multiple GPS devices wouldn't have tested some of them to see how accurate they actually are at measuring distance before writing that article. Other than Bob Baumel's tests of his iPhone (which we all agree now is not an accurate GPS device for measuring distance), I haven't seen anyone report tests of a GPS device measuring distance that have not shown them to be accurate.