How about a lower bar, then: has blockchain/crypto tech ever been demonstrated to be economically viable for anything besides a store of value? There has been a ton of hype about e.g. Cryptokitties but it seems to follow a hype and bust cycle every single time.

How about borrowing and lending using MakerDAO and Compound? Or, a decentralized prediction market through Augur? Or, decentralized derivatives through dYdX? Once the Oracle Problem is solved, you'll be able to trustlessly buy synthetic decentralized shares of the S&P500 without using a brokerage. Some people have already done this as a proof of concept.

Let me have a go then:

1. A way for businesses that cannot get a bank account to accept digital payments. Think porn sites or marijuana stores.

2. It's used on darknet markets for obvious reasons.

3. You can use it to pay for things even without a bank account.

4. Send money to anyone from anywhere.

5. It can be used as a decentralized timestamping service[0].

6. Lower fees than other digital payments (credit cards take 1-4% fee per transaction, plus other costs).

Cryptokitties is just like Beanie Babies and in my opinion not very interesting.

[0]: https://whycryptocurrencies.com/timestamping_service.html