I think decentralisation is the wrong way to look at it. The protocol itself becomes the point of centralisation with many providers implementing in federation. Nothing to stop any of them being commercial (they may have all kinds of ease of use and features on top) or not. Since you lose the network effect of bigger player owning 'footfall' (e.g. ebay), there should be less natural tendency for monopoly and greater incentive towards user needs. For example, a more privacy conscious provider may choose to only share limited data with the wider federation. Others may be all about the exposure and getting more followers. (I'd expect federation to allow users a way to export their full profile+history and migrate to another provider taking their data away with them.)
If the protocol is what makes it work, then it's definitely not centralized. Look at email, for example. I love it, but it's past its prime, because it just doesn't evolve to meet new needs and new mental models.
Contrast this with how well proprietary messaging platforms have been taken up and improved iteratively. Or even look at how Google is advancing GMail. Email is nominally federated, but they have circa a billion users. 44% of US adults report using it, and it's 61% of 18-29 age users.
Standardisation would have been a better choice of words.
But this sounds like the old proprietary vs open source arguments. 1. The biggest tech companies in the world have struggled to keep up with the pace of innovation in open source. 2. Scale of userbase counts but standards facilitate this by creating a network and market. 3. There is a compelling business case for having an ownership stake in your tech platform which revolves around roadmap, security, and longevity. Longevity is counterintuitive but FB can bite the dust like Myspace, and Google can go the way of Woolworths. Huge institutional brands can evaporate fast and barely leave a trace; many young people today have never heard of Madonna, David Bowie, Led Zeppelin, unthinkable to older generations.
The way to differentiate on standards is via features and service, which gmail does well (spam was the killer feature which got most using it). But just lately Google has started looking at proprietary extensions to web and mail, and it's like Microsoft in the early web all over again. We saw how that turned out, even though it was unthinkable that MS wouldn't "win the web" at the time by their sheer scale, brand recognition, and a market share which amounted to a de facto monopoly.
(And funnily enough, Google seems to make the same product mistakes as MS, where it tries to boil the ocean in an orgy of featuritis. With a subset of the tech behind Wave they could have made a great competitor to Wechat/Line etc. Look at Slack - it costs around the same, possibly more than, Office 365 or Google Apps for Business, yet does a tiny fraction. Everybody knows Slack's only real differentiator is UX and service e.g. not having to run chat servers, bouncers, search etc. Again Google could have pieced a lot of Slack together from existing components in Wave. Feels like they are very shaky when it comes to executing product and it could be their undoing.)
