A later chart in this same report shows an interesting phenomenon: in the previous to the weeks with build volatility, there were an abnormally high number of file commits. Could it be the case that the changes committed to the system during this week created almost three weeks of instability afterwards?