the main issue is that the standards to enable this vision are still in their infancy and most people seem to be unaware that these even exist.

Decentralised Identities (DD of https://identity.foundation/) together with Verifiable Credentials (VC from https://www.w3.org/2017/vc/WG/) are the base premise for giving control to the user over their data.

But there is no standard for encoding the relationships between people and organisations. There is no standardised VC schema which allows me to encode that I'm the owner of an organisation.

On top of that, identity of organisations is being created (https://www.gleif.org/en/) but not digitized yet (we're working on it). The user or machine does not have an easy way to authenticate the organisation and verify that they're interacting with an authorised agent of an organisation.

On top of that, even if you could standardise and everyone agreed on the standards (like TCP/IP for networking) very few people understand what goes on within an ML algorithm and what the output actually means.

My point is: I agree with your analysis but think that this vision is still some years away