I don't think "keep very close tabs on" can prevent an attack on an account before it happens. All it can do is help clean up the damage quickly. Which is what happened here. The offending tweets were deleted in 10-15 minutes. That's a pretty good response time.
What could prevent an attack is limiting the features available to high-value users. One could imagine, for example, limiting 3rd party API access. Depending on what actually happened here, that might have prevented this. But there is a downside of reduced functionality for the owners of the accounts in question. There's always tradeoffs.