I don't believe that hypothesis is compatible with the available evidence.
As a counterexample, I offer the military industry. Under your hypothesis, the all-male military would have remained as such forever, because women would have seen the near-0% proportion of females in the military, and would never have attempted to join it. So how did the military get up to about 15% women?
Clearly, it was because at some point a decision was made to actually accept those who attempted to enlist, and not because the industry already had enough women in it to encourage the attempts to be made.
Nevertheless, it is still a perfectly testable hypothesis, and an experiment might still be in order before we go around saying it's no good.