This is a long and chewy article, and I have only read about half of it and skimmed the rest.

I don't think I'm going to disagree with any of your logic or conclusions, but your terminology is a little inconsistent, and I'm here to not only shore it up, but also, to show you a little more about what the specific badness part of blockchain is, and at the end, present a challenge to web3 advocates!

----

The thing you call a "centralised blockchain" doesn't really exist as a computer science concept.

What you mean is a cryptographically signed ledger - a list of transactions that cannot be altered. You can even distribute that ledger, using a Merkle tree - patented in 1979.

So what you're talking about is not a "centralized blockchain". It's a cryptographically signed ledger, not a blockchain at all, because it does not solve the Consensus Problem.

You can look at the Git ecosystem, all interoperable, with multiple huge providers, local providers, down to people such as myself with a couple of hundred git repositories on my disk, to see a huge distributed cryptographic ledger working effortlessly.

No, the blockchain is the tool that solves the Byzantine Consensus Problem. Only an untrusted, consensus distributed cryptographic ledger is a blockchain!

And yes, in computer science terms, the blockchain a breakthrough - it solves a very hard problem suspected for a while of being insolvable. Amazing! Wow!

But that power comes at an incredible cost in resources and performance. And so far, there has only been one application for the Byzantine Consensus Problem, which is completely trust-free, depersonalized currency - literally luftgeld.

Which you seem to explain in more detail to come.

I do want to add one key point - it's that so far, every single web3 concept that has been trotted out that isn't "trading cryptocurrencies" is something that could be done with a boring old 1970's technology distributed cryptographic ledger for literally 0.1% of the total CPU, memory and storage cost.

So I present to you The Git Challenge.

Give me your web3 end-user idea, one that that isn't about moving cryptocurrencies around and I'll sketch a design that will do exactly the same thing but better, using literally a few scripts on top of Git and for 1% of the resources, and be able to deal with things like fraud, mistakes, legal action or agreed-upon revokes.