The thing that changed was the realization that there were significant numbers of asymptomatic carriers. This was not something that was fully appreciated at the initial stages of the pandemic response (in the US at least).
Masks really don't help protect you as an individual. There isn't a lot of data available, but it has been studied (Ars has a good overview here [0]). The main take home message from most of the studies was that infection rates were largely unchanged or worse when people wore masks -- one theory being that there was a false sense of security. Contaminated air will simply get past any non-medical grade N95 mask (and N95 masks that aren't properly fit won't protect from viruses). So telling people that masks didn't help was completely accurate.
But, when you have asymptomatic carriers, the masks help protect other people from you. They do this by limiting the distance contaminated air travels from your mouth/nose. So, masks are only effective when they are used by everyone -- and in the context of asymptomatic carriers. Also when people are able to be distant from each other, masks really don't help much more. It's when people can't be distant that they help -- but again, they aren't meant to protect the wearer.
The health authorities (and the media) in my Midwestern city have been very clear about the motivations for masks. Unfortunately, communicating this reasoning has not been as effective. People tend to be more concerned about protecting themselves than the others. (Which is understandable)
[0] https://arstechnica.com/science/2020/04/should-you-wear-a-fa...