"Web 3.0" isn't the metaverse. It was the Semantic Web.
Web 3.0 was designed to give individual web publishers and consumers the capabilities that giants like Facebook and Google have.
Documents had structured data, marked up in rich RDF or Owl ontologies that could expose APIs for your data. You're probably familiar with RSS and Atom and the wonderful ecosystem of tools they had, but there were hundreds of RDF schemas. FOAF, for example, would let you publish and consume contact info, and even encode friends lists if you were interested. There were lots of these.
This was also at the time of peak Bittorrent/P2P, and we were really close to having schemas for sharing news and commentary in a P2P swarm. Imagine fully decentralized Reddit and Twitter (not federated!), where identities were anonymous by default, comments and upvotes were cryptographically signed and published, and you could curate your own peer group, interest graph, attention algorithm, and filtering. We were almost there.
Clients that operated on RDF schemas were also faster and less cludgy than HTML. We could write native code and GUIs to consume and publish this data, and we were close to evolving the HTML soup into something better.
Unfortunately, this is right at the time Google and Facebook had grown immensely from VC and ad money jet fuel. Academics and builders were distracted. People flocked to centralized platforms because it was easy and our tools were immature.
We were really close, but the tech cos got there first, got lock in, and then started messing with the commons.
Please don't forget that "web 3.0" wasn't crypto. It was more than about payment. It was a stack that gave you and I capabilities and potential that the giants have. It was a distributed knowledge graph that would have been a nirvana of innovation and sharing. But we got sidetracked, out engineered, and roadblocked.
Web 3.0 = Semantic Web.
We can still remember what it aspired to be, and we can still build it.
Never forget.