I won't attempt to enumerate the ways that crypto is useful. But i'll tell you how I want to use it.
But first, it's important to understand this. The actual foundations for what crypto can enable is still being built, so if you're expecting me to show you some "killer app"... i'm not going to show you a working version today. Is that because Crypto is vaporware? No, i'm seeing important stuff deployed all the time. It's because the underlying infrastructure is being built as we speak. One example, today ETH transactions are expensive. I bought an ENS domain (which hey, that's cool) the domain cost $5, but the gas cost $200. That's a problem. Fortunately, that's a problem people are working on. ETH is building out L2. AVAX (the chain I am most heavily invested in) is using something called subnets. A subnet for example allows you to have a custom VM, and isolate chain traffic (gas cost is a function of supply and demand for blockspace, if you both increase supply of blockspace, and decrease demand for specific blockspace you can reduce gas prices). The early examples of subnets are demonstrating this theory is true.
With cheaper transactions, and custom VM's the application space will widen. On my preferred chain AVAX, that's stuff that is JUST now being released (like first one was deployed a few weeks ago). The stuff that is becoming popular today (Defi, NFT's etc) are foundational components.
So with that laid out, here's my use case. I want to run a SaaS. How is a blockchain superior to "traditional" methods? A saas managed by a DAO has a central advantage. No employees. As a corporation i'm limited by the resources I have employed for me. Need a feature? Let me prioritize it, add it to a sprint etc. Corporations have employees, and they can only move as fast as those employees allow it. A DAO with zero employees has the entire internet as a potential contribution source. But unlike open source development, it's not done by "good will". You can pay people for their contributions in a trustless way that allows ANYONE to contribute and earn. Also, and this IS a benefit, in a future of GPT3 and AI assisted coding. There's probably going to be swarms of bots searching the space for work. Posting work, providing a way to pick it up, and integrate it into your project is going to be WAY faster in the blockchain space, than what corps will be able to acheive. But there's a few missing pieces before this can work. We have services like Akash, so throwing services into the blockchain is possible today. But managing a repository on the chain is not. I was contemplating using stateful precompiles in a Subnet VM to do this. The next problem is bridging SUCKS, but all these raw components are scattered around. That makes it difficult, even with the pieces defined to be combined. Solvable problems, but things that need to be solved.