I mean, to play devils advocate, 5 hours is a pretty short amount of time between the tweets and the release of a video.

Oh yeah, that's never been the point tho. I totally agree that it takes a whole lot more professionalism than what they ever even conceived existed, to notice a former employee tweet-storming problematic matters about you, in the middle of a crisis.

With proper processes in place, however, it is all automated, and very simple: you either have an event-based, or a poll-based automaton, checking all your former employee public social media feeds and websites for potentially problematic leaks, infringements and information in general. With triggers such as company name(s), brand name(s), internal project names, management staff first and last names, and hot terms like "fired", "I quit", "abus*", etc etc.