All of those are interesting properties, but they are all unproven at scale and it's unclear whether they are "better" for humans than the things they claim to replace.

- Smart contracts can only effectively secure digital assets and entitlements that can be transferred electronically. Ultimately anything physical requires legal and physical entities to enforce. As an example: it's fine to put property contracts on a blockchain, but proving who owns what property is rarely a real world problem. We have land registries for this, and it's in everyone's interest to keep copies of sales contracts and sign them under witness. You still need bailiffs and lawyers to get squatters or bad tenants out of your property.

- I don't think 'centralized' exchanges are really a problem. There are hundreds of stock exchanges throughout the globe, tens of thousands of brokers and huge off-book markets. We live in a highly decentralized system already. There's an argument that Average Joe can't access these markets, but this is changing with traditional solutions already (see: Robinhood etc).

- decentralized finance (lending, investing, etc). Again, this doesn't work for lending secured against hard assets like property (i.e. mortgages). You need courts to deal with grievances over ownership, and this is where the costs lie. Solving the paperwork problem is not interesting.

- instant borderless value transfer. Well, Transferwise and many more obscure, specialized companies offer this already, and waiting a 24 hours for a big, global transfer in the worst case isn't really a big deal.

The bottom line is the needs of most people are already well served by traditional blockchainless services, and where they're not it's not obvious that they can't be.