One other explanation which I have seen in practice on my images is wind vibration of the camera. This can have a negligible effect on the star images since they are integrated over minutes but show up very clearly on meteors which are only present for a fraction of a second. You even see this on trailed objects like minor planets in long exposure images. If this were a real, physical effect on the meteroid itself you should be able to calculate the transverse acceleration using the amplitude of the corkscrew and the presumed range and along-track angular rate. It would be interesting to see whether this looked physically reasonable given the kind of off-track lift force that you might imagine.
British Astronomical Association
Supporting amateur astronomers since 1890