This article however is a really good counterpoint - it lays out the most reasonable case for Bitcoin as a new class of asset which is useful as a representation of value for distributed applications. This is the strongest case for blockchains I have seen.
It also points out the severe deficiencies inherent in a distributed model which make it unattractive on many measures compared to existing solutions, which for some reason current 'crypto' (sorry cryptographers) advocates find it incredibly hard to do - perhaps because they have come to value the asset more than the use case. Decentralisation makes bitcoin slow.
The only bit where it breaks down is in its insistence that 'Censorship resistance' is a fundamental advantage of distributed applications. This appears to assume that the network will not be policed, and the endpoints cannot be policed. Your computer is vulnerable, your servers are vulnerable, the bitcoin developers are vulnerable, exchanges are vulnerable, miners are vulnerable. All of those pinch points can and will be bribed, cajoled, shut down or worse coopted if significant real value is captured in distributed networks.
no one can stop me from using Bitcoin to pay for something
Why yes, yes they can :) Or even worse they can turn bitcoin into something that you wouldn't want to pay with or are forced to pay with like company scrip. Exhibit 1, the cryptoruble:
https://cointelegraph.com/news/breaking-russia-issuing-crypt...