Oh, ok. I didn't know that. I thought they were terming it that in preparation of quarantine, immunization and possible destruction of all but heavily guarded laboratory samples as the goal would be for disease contagion.
Well, it would be interesting if such movements turned out to be heavily dependent on particulars of social media, but I personally believe they are more robust.
While Twitter, Facebook, etc are useful tools for protest movements and other social dissent, but I don't see the specific details of them (which a study like this would be looking at) as shaping the movements so much as amplifying them. If it turned out that something simple like trending on Twitter had a substantial impact then that might be something that could be used as a sort of immunisation (change Twitter's algorithms to prevent such trends, or at least have them vanish faster) but I doubt that's the case.
But trying to see how trends online relate to real life events could be interesting. If we know that
these tweets directly relate to
this real life action (which has been retroactively studied and understood in it's scope and form) then in the future we might better understand
this ongoing action from
these new tweets. Do the tweets from a protest about to become violent have any particular shape? Does a thousand person rally with good PR look different online to a ten-thousand person rally without? Could be a lot of interesting data there.