The cost of regulation is a decrease in both convenience and innovation. By eliminating regulation, you get a whole host of new startups that were previously held back, some of which solve genuine problems that have no existing solution. Consumers flock to these startups because right now, in this moment, they solve problems better and are more responsive to customers than the existing regulated incumbents.
Many regulations solve problems that only appear at scale, so as long as the new startups are small and voluntary, regulators take a hands-off approach and let these startups enjoy their competitive advantage. It takes time for regulators to catch up, so for several years, these new solutions can grow and get new adopters. Eventually all the bad behavior that caused the regulations in the first place appears, and there're calls for regulation, and the new boss starts to look an awful lot like the old boss. But people don't make their purchasing decisions based on what's going to happen in 20 years, they make their purchasing decisions based on what they need now.
You see this with a lot of dot-com era startups. People knew in 1997 that Amazon was going for monopoly and was just going to jack up prices when they achieved it; hell, Jeff Bezos even told investors as such. But consumers didn't care: we wanted convenience and low prices now, and even if we did without, other people would give Amazon their business, and all we'd succeed at is disadvantaging ourselves. Similar with Facebook; most people knew they were trading away their privacy (Zuckerburg's "dumb fucks" IM was made public in 2010, and he said it in 2004), but goddamnit, people wanted to see what their grandkids were up to.
It may end up rebuilding the current regulated banking system. But if the regulated banking system and the unregulated banking system are sitting next to each other using tokens that are interchangeable with smart contracts, that seems like it could be useful.
Also proof-of-stake has been promised for so long I would be embarrassed to talk about it if I worked at the Ethereum Foundation. Delivering things that work isn't their strong suit.
Division of tasks, and specialization naturally trend to hierarchical organization for the same reason divide and conquer algorithms are so efficient. Separation of concerns is powerful.
There is much different of 5-6 shoemakers picking the same person to handle their finances so they can focus on making shoes. But, how many shoe makers can offload their finances until you have a bank?
