I got a little too excited at first about this paper about airborne transmission and masks. pnas.org/content/early/… See the thread for more /1
Fig. 2 shows cases leveling off exactly when masks are required, but there should be a delay. Is this actually when effects of stay-at-home and distancing orders from weeks ago kick in? /2 
In part A of this figure, I feel like I could just as easily draw a straight line from the stay-at-home date to May 10 and get an equally good fit. Again, effect of masks should not kick in immediately. /3
Fig. 5: Unfortunately, the shield looks kind of like a mask. Social distancing should also attenuate short-range airborne transmission, so 2nd purple arrow under A should be short and dashed. In C, masks and distancing should be shown in parallel, not series. /4
As pointed out by @GMMeyer and others, many in NYC were wearing masks before the order. Not sure if this helps or hurts the case. The challenge is that gov't orders do not produce a clean step-change in behavior. /5
Humidity discussion needs to consider indoor conditions. Coagulation and growth indoors are unlikely. /6
There's probably something here, but it may not be as clean as portrayed in this paper. Inclusion of experts in public health and policy would have been helpful. /7
