one cannot in good conscious claim they have "data ownership"
This is a misconception, an assumed evolution of Web3.0 in your argument.
It is true that blockchain aren't that innovative in the Web3.0 sense and platforms like Technorati back in the day were the real innovators (also microformats and schema.org by Google).
But at the same time blockchain are the platforms trying to bring the semantic functionality to the user. Because that semantic functionality is key for Web3.0.
And that is perfectly visible in concepts like MAKER DeFi or even here on Hive. The data can be accessed and used by different bapps (Blockchain apps, most are centralized rather than decentralized).
As such the argument about Web3.0 should be built around data portability rather than ownership. And there decentralized blockchain do play a rather nice role and as we will see always more interaction between chains they will become a true multitude of Web3.0 nodes.
Blockchain are the "Fat Protocols" (see placeholder.vc) which are the nodes in Web3.0.bapps are users.
I totally agree with your point of lack of data ownership btw. A lot of bapps also run Google Analytics and many other centralized scripts which... is kinda ironic and also sad at the same time.
Lastly, while the outcome was exactly what you sketched, Web2.0 was defined by User Generated Content (UGC), not by offering the account layer as such. Self-hosted blogs are also Web2.0. But I realize that's semantic pettiness since the outcome was exactly what you described. I was always a a of openID, the first large scale effort for "account portability" but once the Goliaths understood they could centralize said portability by using their own vast database of accounts and making it faux-portable it was game over for openID.
Semantic functionality is indeed the important distinction of the original Web 3.0 idea, but a web of independent databases that contain different states from each other is much richer than a web of the same database distributed across a network. Database technology is key to moving in the right direction and blockchains are an innovation in the space and the ability to query information in them based on certain relationships is improving, but they are limited in their global scope.
It is encouraging that people are starting to utilize information in these database and organize these databases in interesting ways to run applications, but at the same time, because of the lack of diversity between databases, I feel that the potential and promise of a semantic web isn't quite there yet. As the way applications run now, there is still a central application running from a single computer or a few computers, as opposed all machines accessing data as running application functions that allow some degree of flexibility depending on the user or the underlying data being queried from the database(s).
That flexibility I feel is the biggest promise of semantic technologies, we got a taste of it in early Web2.0, but it was phased out because uniformity is more efficient and cost-effective.
Needs be noted several blockchains chose to embrace "Web3.0" as a term and that's a good thing because it gives them an actual purpose, a raison d'être so to say. But Web3.0 never demanded anything such as decentralization and preceded even BTC by years (IIRC Sir Tim first mentioned it in 2003 or was it 2005?)
Yet,they are harmed, have no nimbleness in their evolution where centralized organisations can and do move much faster. Blockchains are Apache.org style slow squared. Both held back in pace and actual conceptual evolution by their governance model.
But we are getting there. Maybe not yet to the point you think web3.0 should decentralize; that would have us look at the original torrent system, ipfs already does well in that aspect but hasn't yet hit mainstream (momentum is consolidating towards further use of the protocol by other chains). Slowly but surely, it will happen and I don't really expect any Goliath to enter the game and hijack that functionality. Even if they could design the "Fattest Protocol of All".
Btw I did enjoy this angle to creating discussion. There's a lot of truth in it.