The study does acknowledge finding some fibers which "had the spectral characteristics of cellulose, compatible with cotton fibers" (if it's cotton, why not just say cotton?), but this doesn't look (or move!) like any cotton I've ever seen. Note the increased activity at around the 8 min/35 sec mark, as the music on my TV becomes louder:
The twisting and bending movements are not the result of a passing breeze. The way the fibers move can be described as both robotic and serpent-like. I believe this fiber was particulary active due to high humidity, as I have noticed the fibers are much less motile when dry.