Yet somehow you didn't find even the Trek "research" in your "several hours researching."
FWIW, the Madsen, Andersen (
AFAIK, no relation to Hans) and Lahrmann published research was on
*flashing* running lights, and found a ~50% reduction in multi-party crashes. They did not study the difference between flashing and steady running lights, since at the time the induction lights used only supported flashing.
A followup study on the safety impact of
a yellow bicycle jacket was done by
Lahrmann, Spaabaek,
Madsen & Ossman, with similar methodology. Also finding approximately ~50% reduction in multi-party crashes.
So, the existing research covers annoying people on bicycles, and covers people wearing clownsuits on bicycles.
Future research no doubt will cover annoying people wearing clownsuits on bicycles. (It is unlikely the combined effect is completely additive.)
-mr. bill