A paper argues that most decentralised use cases are not explored because we are focusing on the wrong spots

Stumbled upon this paper while chatting at an event in Lisbon. After giving it a read, I found the idea pretty interesting. It says that decentralization isn't about replicating Facebook with an equivalent dapp but about enabling stuff that was previously impossible if we were stuck with the likes of Facebook. This perspective got me thinking, and I'm curious about how others see this. Do you think there's merit in looking at decentralization from this angle? We are probably like 2016 again so it's time to think.