Social media moderation has to balance "engagement" with the potential for bad PR or liability for the company. It turns out that content that is against the rules is also the one that generates the most engagement, so enforcing the rules as-is is bad for the bottom-line.

Thus for every piece of content that is potentially against the rules, the actual condition for removing it is whether the expected engagement potential outweighs the probability of someone rich/well-connected getting inconvenienced by it and how much inconvenience would it be. Content is only removed when the liability potential exceeds the profit potential.

At the beginning the reports were ignored because the system determined it is more profitable to stay up. I'm not sure what "his pleas to take it down" refers to, it would've likely been just his staff members flagging it with their personal accounts and those flags having very little weight. Eventually either someone managed to talk to a human and/or a letter to their legal department arrived, or the content achieving enough impressions to become a risk which caused the earlier flags to actually get reviewed by a competent human, at which point they realized what their liability was and quickly removed it.

You should expect to see an apology from their PR department soon and a promise they'll do better next time.