Why?
The tool in question is likely used by low level support/abuse control workers. The huge pressure put on social media firms by liberals in recent years to crack down on "abuse", "hate" etc means they need a vast army of people to review complaints about harassment, "fake news", account hijacking etc. Those employees aren't all sitting in expensive San Francisco on a corp VPN, are they? They're probably going to be in places like India.
From the mention of BeyondCorp, it feels like there are a lot of Googlers in this thread who aren't really familiar with how Google handled the same problem, or at least, used to. For example back when Orkut was big there were huge numbers of people in Brazil who had the power to censor content, ban users, handle victims of phishing and so on. It was the only way to scale the moderation users and governments there demanded.
An ideal user admin tool is very fine grained. But once account hijacking entered the picture, it gets hard to truly restrict takeover permissions to a tiny number of people, because accounts are constantly being taken over by third parties and need to be reset back to the true owner via manual intervention. Attempts to automatically handle that are very hard, I know from experience. Hackers like to abuse any system put in place to stop them taking over accounts (like 2FA) to stop the true owner taking it back once captured.