So the argument starts with a (rather bizarre) claim that blockchains are centralized because one have to use the protocol that the rest of the network is using. The argument goes on with an attempt to contrast blockchains with the "web as it exists" and claim that one can "speak whatever protocol you want" in it -- with a caveat that "one have to put significant work" to convince others to use it.
It is quite ironic that author is basically explains the concept of a hard fork [1] while never even mentioning those. Most likely because he is not familiar with basic blockchain concepts. Or, to put it more bluntly, has no idea what he is talking about.
I agree that there's a little confusion in the article regarding the role of protocols in the decentralization equation. That said, I do think the larger point (global consensus protocols are just another form of centralization) is valid and worth discussing.
No - that's not a "little confusion". It is almost total lack of knowledge of the subject at hand. So many concrete examples by the author are not valid. Changes and forks in blockchain protocols happen all the time. You can start your own local blockchain in your own private IP subnet or whatever. You can create your own private ledger on the global net in less than a minute [1].
You could say that this new form of centralisation has the potential to allow governance systems that breakaway from, modify or usurp nation States. All of it being economically incentivised. This is pretty interesting IMO.
It is quite ironic that author is basically explains the concept of a hard fork [1] while never even mentioning those. Most likely because he is not familiar with basic blockchain concepts. Or, to put it more bluntly, has no idea what he is talking about.
[1]: https://en.wikipedia.org/wiki/Fork_(blockchain)