Data is presented that shows a certain trait or tendency- in this case drivers tend to give unhelmeted riders and female riders more berth- and then the researcher gives his/her guess why this is so. The data is the important part and the researcher's best guess is not part of the data. You say at the end his analytical skills are lacking but what you've been attacking the whole time is the expirement itself. I'm not defending his guess, his guess means nothing to me, I'm saying it sounds like he conducted the expirement correctly and therefore the data has meaning.
I think that you should re-read my critique of what we know of the experiment. Obviously the article is not a comprehensive analysis of the experiment.
That said, you again, as he appears to have done, appeal to a sort-of-psychological account of drivers' reactions to helmet-wearers.
THE STUDY HAS NOT CONTROLLED FOR DRIVERS, and I think that we have all agreed that it really can't control them. Appealing to them as an explanation of the data is bogus, regardless of how the data was collected.
From the perspective of "the scientific method" (I've done my share of experimental physics), there are problems with the analysis. You shouldn't go using uncontrolled factors as causal explanations -- you can't reproduce the behavior.
From the perspective of a logician (I pursued philosophy in my later academic life), there are problems in that plenty of counter-explanations deal as well with the collected data and those alternative explanations have not been addressed.