It actually seems quite high, assuming participation follows a power law [1] distribution curve.
Take, for example, the recurrence of a given individual's contributions to open source projects - many have not contributed, but I'd bet quite a few only do so once.