Originally Posted by Brian Ratliff

I am stretching somewhat, but the only way I could imagine him riding on a sidepath and running into right hooks and chroming out into a headon situation is that he was riding fast (~20mph) and he never yielded at an intersection to cross traffic. He rode in a manner which is commonly used on the street, but is suicidal on a sidepath.

This is not an apples to apples comparison which can arrive at a risk ratio. He simply applied his on-road cycling parameters to the sidepath environment and found that they don't work (which is no real surprise). This is all the test shows. It cannot be extrapolated to form any sort of absolute safety comparison.