There have recently been highly creative ID deception and theft attempts intercepted. In one, the perpetrator took over existing social media accounts of family and friends (that were otherwise abandoned), either by forcing them or by buying them. In another one, the perpetrator created social media accounts for their victim. These two examples involve real persons. In many other cases, fictional profiles are created based on individuals. In even more cases, the perpetrator claims to act on behalf of an individual. This is most common with accounts they claimed are 'managed'. The victims normally have no idea what's going on or don't understand their identities have been monetized. Hivewatchers deals with all these but even then its hard.

Related, to your point of identity verification, I recently attempted a different approach at Hivewatchers with an account that was accused of being a fake ID. What I did was go through all the messages sent by the person. When reading them back to back, you see that on one day the person is happy and answers one way. On a different day, the person is upset and answers another way. They will give different answers to the same question based on what's going on. A fake account or a thematic account (a type of 'managed' account) can't do that. It will always respond in a similar sort of way because the person is acting. It will have some variation, but it will be minor and playing to a template. Me revealing this here won't give anyone the upper hand in defeating the ID verification procedure; its beyond their acting capability.

As you know, in general, there is a lot of misinformation on Hive. People who spread it are interested in spreading it because they build up their fan base off it and create an echo-chamber. Everything is misread and twisted. We're used to seeing it in general but it's particularly detrimental when it's about Hive itself, as it affects the entire ecosystem. For example, that there's some mythical secret group deciding which proposals get funded. Or that the top witnesses are colluding. Due to the sheer number of these people, using crowd-sourcing to gauge the credibility of information while drawing on the same groups that have been misinformed won't work. Many "confirmed facts" aren't based on experience or education, just guesses. People make observations, they grow to believe them, others confirm their ideas, and now there's a majority view.

What I'm getting at is that a system that tests for variance in content, wording, subject matter and range of use by an account could be the start of a good web of trust. A real person of trust will have an opinion that changes, they will be different on this day from the previous, they will try different dapps, they will talk about things in their life and grow as a person. A fake one won't.