Some of this is probably driven by mobile usage and unifying the experience between mobile <> desktop.
But the truth is a team almost certainly tested this and measured an improvement of some topline performance metric.
(Hacker News articles comparing YT before and after screenshots is not one of their topline metrics.)