Disclaimer - If you haven't read my disclaimer yet, make sure you do so here. TL;DR version - Buyer beware, I am not an expert, I am fumbling my way through this like the rest of you.
Also, I hold a little bit of Bitcoin and Ethereum.
The promise certainly sounds good
Blockchain promises us something absolutely amazing - the decentralization of everything:
- Whereas all currencies today are run by central governments, Bitcoin is not controlled by anybody because it is decentralized and therefore nobody can manipulate it
- Whereas platforms are built on private corporations' servers, Ethereum is not controlled by anybody because it is a decentralized platform that nobody can manipulate
- Whereas all of our data is controlled by a handful of big companies, files stored on the blockchain are distributed everywhere so nobody can control it
This promise of decentralization is key to the entire value proposition of blockchain.
But tech has touted the benefits of decentralization before, and it didn't live up to the hype.
The idea behind networks and the network effect is that you have a bunch of nodes that are more-or-less evenly connected, so that everyone can learn from everyone else. This is the grand bargain behind social networks - by sharing information, the best and most correct ideas will bubble to the top while the wrong and worst ideas will fade away. Information wants to be free - and shared - and therefore reason and logic will win the day because of the free flow of information. Good ideas survive (eventually), bad ideas are disproven (eventually) .
That's what we were promised.
The reality looks way different
But it's not how it turned out. Social networking, far from having a unifying affect on people, has turned out to have the opposite effect. It's made us more divided than ever. And, far from weeding out the bad ideas, it turns out that wrong ideas flourish in social networks.
The reason is (partly) because in a social network, nodes are not evenly distributed where information flows freely amongst everyone. Instead, it turns out that there are bottlenecks everywhere. Groups are localized among each other, and connected via super-connectors (pictured in blue in the diagram below). We stay by ourselves in our groups and only occasionally interact with the outside world, and it's done via the super-connectors.
This ends up suppressing the free flow of information, and reinforcing our own particular points of view, since information depends upon whether or not they can get through the chokepoints.
The tech industry seems baffled about why this occurred. Wasn't information supposed to be free? Didn't it want to be free? Wasn't free expression and debate supposed to sift out the bad ideas? How could this be? 
Yet this was entirely foreseeable. The decentralization of information has occurred before, the Internet is not the first time. The most obvious example is the Protestant Reformation. Whereas Christian theology used to be centralized in the Catholic church, the Protestant Reformation's rallying cry was that all people could be their own priests, and everyone could interpret holy scripture for themselves. The interpretation of divine revelation was meant to be free, not centralized.
While the Reformation succeeded in reforming the Church, it also caused a lot of divisiveness, wars of religion, and tens of thousands of splinter denominations. That's what happens when you decentralize, anyone can come up with a particular view; and then if you have enough skillz to make a particular outlook popular, you can get people to join along with you and embiggen your group. I suspect Martin Luther did not foresee that people, upon interpreting scripture for themselves, would come up with such different interpretations and would form splinter groups as much as they did.
In the same way, tech did not foresee how the decentralization of information on the Internet would lead to the insulation of people seeking to reinforce their own particular viewpoints and biases, rather than seeking diversity of opinion in a quest for truth. Yes, that exists. I do it, you do it, we all do it. But we also prefer to reinforce our own viewpoints in many areas of our lives.
As a representative of the tech industry who was caught off guard by this, I plead guilty. I don't know whether or not I should have known better back then... but I certainly do now. The results of decentralization can be mixed .
What happens when we apply it to new technology?
Which brings me back to Blockchain. As I said earlier, its biggest premise is decentralization. Yet decentralization is a mixed bag. Will it deliver on its promises? Or will it fall flat? Or worse, will the results be mixed such that we become so dependent on the good that we can't roll back the bad, lest we simultaneously rollback the good?
Bitcoin has been a roller coaster since its inception. It's up bigly since it was created but has undergone massive corrections of 30-60% every few months (it recently corrected 45% from its peak this past December 2017, and only took a week to do it). Is this massive volatility a temporary thing until Bitcoin stabilizes as it gets more adoption? Or, is volatility a feature of Bitcoin because of its decentralized nature, where anyone can come in and buy/sell with no financial oversight?
All this volatility is not a good thing; between buyers and sellers, if the medium of exchange is likely to fluctuate wildly, then what you pay is likely to be far different than what you agreed upon. That makes it less likely you will want to use it as a medium of exchange which was one of Bitcoin's original benefits - to act as a decentralized currency (although as I have said before, it behaves more as a store-of-value since the IRS treats the dispersion of your Bitcoins as a taxable event).
Thus, when it comes to using blockchain as a currency, decentralization has the benefits of being a deflationary currency that is resistant to manipulation by central banks, but it's still potentially vulnerable to manipulation by the millions of end users who have their own ideas of what digital currency is, and what it should be, and then shunt it off from its original purpose.
I don't know whether or not that's a good thing.
What about decentralized applications? Those act as a currency (sort of) and as a platform to build stuff on. I don't know about those, maybe they are useful and maybe they aren't. For sure, blockchain promises to be disruptive. There are some like Ripple and IOTA that solve real problems; Ripple is criticized for being a private implementation of a blockchain, which conceptually it isn't but in practice it kind of is. But even if Ripple is pseudo-private, is that a bad thing (because it undermines the concept of blockchain as a decentralized application) or a good thing (since it means it can't be accosted by "do it yourselfers")?
I think that this is one of the fundamental known-unknowns with blockchain. Decentralization may be awesome.
But then again, it may be yet another false prophet.
 This blog post is inspired by an article I read in Foreign Affairs magazine (I am a subscriber), see The False Prophecy of Hyperconnection - How to survive in a networked age. It's behind a paywall, though.
 For more on the Reformation, see Majority believe Reformation was divisive, but justified