I did some stats on the effect of the number of contributors after Emacs switched to git last May. I think I summed it up as “meh”, but that was only after a handful of months, after all.
It’s now been more than a year, so I redid the experiment:
(The red line is the changeover point from bzr to git.)
Let’s zoom in on the last few years:
I think my original conclusion stands.