'Dinoflagellate'. It's called a typo.
But he was probably referring to flagellates, which propel themselves via a flagellum. His reason for doing so is to show that a flagellate can spin its flagellum at 1000 RPM, stop in a quarter-turn, then immediately spin the other way at 1000 RPM again. Practically no energy lost. Utter efficiency.
Better than anything we could pull off. Yet it supposedly happened by 'chance'.
I know what a dinoflagellate is, but I wouldn't presume to say what he actually meant. For all I know he was talking about dinosaur tails.