It is certainly a factor that might explain some of our findings in the G&S paper. We did some analyses to look at whether scholars delay posting preprints until after the paper is accepted/published, but we found no difference in posting timelines on average.
