Honestly that's how I figured it worked anyway. Popularity contest. Even the less significant confrences seem to just be platforms where "influencers" (Twitter verified personalities in leading roles at popular tech companies with barely any code in their repos) scratch each others back and swap followers. It was always my understanding that confs were a place for networking with other people just for that purpose. Maybe it's because I'm on the East Coast and all the serious confs are on the west coast.
I believe the author of this comment is referring to industry conferences. Please note these are very different from peer-reviewed academic conferences.
In CS these academic conferences are held in many different locations, and many (most?) top researchers are not on Twitter.
