This is a truly excellent criticism of the state of "crypto" and "web3". As someone who thinks these technologies are interesting, i'm glad someone finally wrote a decent, sincere critique that covered a lot of the very real issues with it.
I think i'd break this piece down into two categories: The first is critiques of current implementations, and the second is critiques of the structural incentives of the technology. I think it's important to separate those things somewhat.
The privacy, security and centralization of Infura/Alchemy are real and important issues, and to a limited extent, derive from the fundamental incentives of the ecosystem. However, what I think critically differentiates "web3" from "web2" is that those platforms are commoditized. Infura and Alchemy are providers of a service that is fundamentally a commodity, they have very little market power. Contrast to comparable web2 platforms like Facebook or Google, who have tremendous market power over consumers who's data they've warehoused. I think this is a really important structural difference between the two. That in no way takes away from the seriousness of the critique of Infura/Alchemy and how they're used, but I think it does somewhat limit the importance of that failure. Anyone can build a new, better gateway platform, and users can switch to it without having to ask anyone's permission to export their data. That's a really big deal.
The more structural critique I think relates to the issue of iteration speed, and the tendency that slow, bureaucratic development processes have to push the technological frontier outside their own scope. I think that's a real, structural problem that any decentralized system has, and its fitting that Moxie should point it out, given that he's famously (and correctly, in my view) resisted exactly these sorts of things for exactly these reasons in Signal (e.g. federation) since forever. I think this critique is the most important and serious critique of the crypto space in general, and if anything is going to bring it down, this is it. This problem remains largely unsolved at this point, but whether or not it can be solved is going to hinge on the quality of the group coordination mechanisms people are able to devise. I'm personally optimistic that these things can get figured out, but they are very hard problems.
An important thing to note here though is that a lot of things actually work just fine with this kind of bureaucratic/slow iteration process. Consider core web protocols like HTTP, SMTP, or even something like x86. These things tend to be "low in the stack", but that is exactly what the underlying crypto infrastructure wants to be as well. The more general your platform, the less quick iteration you require. Whether or not crypto platform are able to deliver something like this remains to be seen, but it is a thing that does happen and works ok in many areas today.
> “It’s early days still” is the most common refrain I see from people in the web3 space when discussing matters like these. In some ways, cryptocurrency’s failure to scale beyond relatively nascent engineering is what makes it possible to consider the days “early,” since objectively it has already been a decade or more.
I'd also like to point out that most prior "generations" of the web took at least a decade to come to fruition. It's easy to forget how long things take to mature, but the fact that crypto doesn't have everything figured out after 10 years is not all that surprising:
https://www.timetoast.com/timelines/history-of-web-20
And this is only going back to the early 90s. Obviously various proto "webs" existed long before then. Very structurally different technologies can take quite a while to sort themselves out, and find their niche.