There are a few interesting ideas with the blockchain, and it was something that was worth looking into at least initially. It's something that could be iterated on, it's not useless. But at this point, it is becoming increasingly obvious to me that as a general strategy, federation is a better approach to decentralization than the blockchains' vision.
Around the parts of decentralization that actually matter to me, and in particular in regards to anti-corporate/anarchist spaces, blockchains have almost universally ended up fairing worse than simpler federated projects. The introduction of a "ledger" in the first place seems to be a magnet for the exact corporate/government attention that people say they want to avoid. I've also seen a lot of attempts at turning the blockchain into a tool for app creation and community management, but as far as I can tell none of them have turned out to be particularly noteworthy or useful -- there's not a lot there where I'm looking at it and thinking, "this shows massive potential."
So when I think about my limited time on this earth, I am much more inclined to spend it following and supporting stuff like Matrix, Mastodon, Peertube; heck, even the really "boring" stuff like just increased pressure on Apple to support web apps better on iOS is going to have a much bigger impact on decentralized, anti-corporate app creation than almost anything I can see coming out of Ethereum right now. It's not that there's zero value in the blockchain as a concept or zero potential, if people want to do engineering in that space then fine, I guess. But I look at what the blockchain spaces are putting out right now, and it's just not comparable to any of the other efforts that are going on with decentralization. It turns out that securely syncing an uncontrolled ledger is not as important of a problem or as useful as we thought it might be.
The closest thing I've seen here to an actually exciting project is Handshake, and Handshake has a ton of problems, it's not a great, permanent solution to the problem of DNS ownership. It just happens to be one of the very few, narrow areas where a shared ledger really does matter.
It's like if you claim the U.S. congress is decentralized because there are representatives from each of the 50 states, which is a valid argument despite congress ultimately becoming a centralized power unit, but in this case each of the representatives just buys more mining equipment compared to their competitors.
TCP/IP doesn't care who you are and where you come from
blockchains solve problems like "how can we order all our transactions and write them on this global immutable, append only log, so that anybody can see them?"
a problem I never had, honestly.
To me less web, and less means imply deeper human investment.
Think about Wikipedia. Yes there is still some bias injected into the editorial process - but is it a better or worse than a truly decentralized free for all?