Dude goes on to say
but my inner statistician/econometrician is screaming you haven't actually established causality yet because you didn't rule out other causes you fucking LIB, you just have a working hypothesis!
Then my inner Marxist tells inner stat nerd to shut the fuck up and think dialectically for once in his fucking life and not go full positivist/empiricist when human behavior is involved.
Well you could compare data from Twitter users vs. non users. If you accept the null hypothesis (there is no difference in COVID incidence between the two groups) that would indicate that another factor (vaccination, better compliance with distancing) is responsible for the drop.