It's just marking trusted people over a trusted person, so no moreso than what you suggest. The goal is to decentralise, so sources of truth should be decentralised, any system used to verify against those sources of truth should do little more than display the data is as concise as way as possible

Community-maintained verification lists may be more reliable - A user could subscribe to something similar to a moderation list, which adds UI to show that x person is marked as real by y lists. I think the transparency make it easy to spot the bad faith actors in this environment.