Your mistake is thinking that the relevant centralization is in the client-server architecture. The real reason that people have centralized social is because it's hard to let people authenticate and find their friends across domains. And even further, there are no good tools comparable to Facebook when it comes to subscriptions, notifications, updates, realtime collaboration, and all those other multi-user things that happen on a centralized domain.
The Web is decentralized, email is decentralized, git is decentralized, bitcoin is decentralized, now what is the analogue for social? That's the reason github is centralized. You get all the goodies of the accounts. Once you disrupt this social layer with something decentralized, you disrupt centralization in a big way. That's the more relevant phenomenon.
Once you have a conversation with multiple parties, you need to elect someone to be the authority on "what happened". Mental Poker is still an unsolved problem for arbitrary conversations and participants. So that may as well be a server. The question is about centralization in the macro sense.
The discussion point of the article is how we can build such technologies for the kinds of things you just listed as needing a centralized domain.
You claimed that they were describing your platform.
But in the 'macro' sense, your platform is nothing like that -- it's a different flavour of Web app server, and fits in with the existing paradigm. Yes it may be part of a decentralised system (the Web), but any large Web application developed using it would be exactly the kind of centralised information silo that the people in the article want to avoid. The article is about how the existing paradigm is insufficiently decentralised.
Your platform could be a great Web app platform, but it does nothing to address any of the points raised in the article.
"It does nothing to address any of the points raised in the article"
Really? It seems that many of the actual quotes from the Fathers of the Internet in the article apply to existing web client-server architecture:
“We hoped everyone would be making their own web sites—turns out people are afraid to.” - Tim Berners-Lee
“People have their friends on Facebook and some photos on Flickr and their colleagues on LinkedIn. All they want to do is share the photos with the colleagues and the friends—and they can’t. Which is really stupid. You either have to tell Flickr about your Facebook friends, or move your photos to Facebook and LinkedIn separately, or build and run a third application to build a bridge between the two.” - Tim Berners-Lee
(by the way, we address this explicitly in our platform)
"Don’t discount Wordpress, it has been embraced by large numbers of people, perhaps the new web should have a decentralized Wordpress type of service." - Brewster Kahle
That doesn't sound like a necessarily serverless architecture to me.
I'm not talking about serverless, I'm talking about decentralised. The article is talking about decentralised. You're talking about decentralised, even.
But your platform is a centralised platform just like most other Web app platforms. It is not at all like what the article is discussing, contrary to your claim.
How would a Facebook or WordPress clone built on your platform be any more decentralised than regular old Facebook or WordPress?
In fact how would it address any of the questions raised?
Is your platform distributed and decentralised?
Is your platform easier for Joe public to publish websites on than wordpress and without a central service provider?
Does your platform prevent government snooping?
Does your platform prevent the accumulation of data in a single information silo?
Does it reinvent the Web, making it, as a whole, inherently less centralised than it already is?
Does it embed key moral principles into the fabric of the Web?
Does it unlock published Web content when copyright expires?
Can it verify financial transactions and authenticate data sources?
Can it do all of the above in a non-centralised way, resilient to network outages, and loss of resources at their original location, across the whole of the Web?
Because those are the things the article describes.
And that doesn't sound at all like your platform, which is another Web application stack, very much the opposite of what is suggested in the article on fact.
Is your platform easier for Joe public to publish websites on than wordpress and without a central service provider?
Yes, just as easily as in multi-person Wordpress. But it does much more than just publish a website. It can power a web app, one that goes in the app store, support social features like contacts, access control, realtime updates and offline notifications, and more.
Does your platform prevent government snooping?
Yes, by preventing data from accumulating in a single information silo. More than that, we believe that access to the global internet (as Facebook was trying to with Internet 2.0 in India) should be unnecessary, as people on cruises or local villages should connect on local area networks, and only access the wider internet when necessary. We believe the older tools in the dialup era were designed more properly than the ones which assume always-on broadband access and only have "isOnline/isOffline" dichotomy.
Does your platform prevent the accumulation of data in a single information silo?
Yes. People choose what organizations they host with. When they visit another domain, they have an instantly personalized experience, with all friends who wanted to share that they also use the service. Much better than "Your friend XYZ is now on Instagram" without their permission, and much better than oAuth. Over time, you might import enough information to have a full-fledged presence in both communities, and choose which one to auth with. Some communities are for your videos, the others are for some group activities or whatever. Any community can embed components from any other community. Any aggregator can subscribe to publishers (with their permission) and get realtime updates, with custom stream types, instead of e.g. Google spidering your site.
Does it reinvent the Web, making it, as a whole, inherently less centralised than it already is?
It builds on top of the web, which is already decentralized, and makes it possible for communities to deploy apps on their domain which rival those on facebook.com, and which allow users to seamlessly use their identities and friends across domains.
Can it do all of the above in a non-centralised way, resilient to network outages, and loss of resources at their original location, across the whole of the Web?
It is designed to power mesh networks such as the ones being set up in various cities. The mobile wireless infrastructure is still centralized, but that will eventually change.
Does it embed key moral principles into the fabric of the Web?
Some of the other things you mentioned aren't really addressed, such as "Can it verify financial transactions and authenticate data sources?" That's more for blockchains. Our platform does not rely on any global resources except possibly DNS.
http://qbix.com/platform