Ah, jes, this has been mentioned before. And there very well may be confounds, such as a tendency of a stereotypical bicyclist to use fewer resources in the first place. Etc... Which is probably true to an extent, but imo the point of that article was that single variable studies are limited.