You may be right. It was only an example, and I'm not an expert on this topic. I was referring to projects like Gridcoin where the computation is used for medicinal research, which seems an useful application. But maybe there are significant problems with this approach.
Another example which could be possible, where the smart contract could also interact directly with the object, is a "proof of connection" contract, which verifies that nodes of a network are online, and rewards those who are. That could be useful for mesh networks, for example.
It's a bit nit-picky but I'm not sure if I'd call decentralized storage systems in themselves smart contracts. I'd define a smart contract as a script that is interpreted by a decentralized network but not intrinsic to it (eg. storage management is an intrinsic functionality of decentralized storage systems whereas LN is not intrinsic to Bitcoin).
At the end, that depends on the implementation (if it's part of the standard feature set or an additional program written on a "smart contract platform"). Bitcoin itself could also be viewed as a "smart contract".
It would be interesting if a Sia-like contract is possible with the current Bitcoin script feature set.
Of course Sia is trying to attract attention by more attractive pricing, but I´m sceptical whether
this will be enough to gain any traction.
I think a Sia-like storage solution, in the long term, should be drastically cheaper than a centralized product. The reason is that it's - at least partly - using unused space, and that there are potentially billions of providers.
Another use case I have been thinking about is related to the hotel industry. E.g. you could
probably automate the booking / check-in process by using a smart contract (a smart contract
checks for availability of the room for the selected dates and releases a code or passphrase
to unlock the room after verifying the successful payment).
For a single hotel, this is another example where you don't need the whole network to calculate your computation because you need a centralized party "managing" it. As a hotel owner, you could simply create that "smart contract" on your own servers - as a "standard software application". Why would you want thousands of nodes to compute that contract?
It would maybe be useful if you're wanting to create something like a "decentralized AirBNB". But here I think the main problem is the trust issue. You as a buyer would have to be able to trust that the code exists, and that you don't get robbed in your "hotel room". It's exactly the "representation vs. direct interaction" issue I mentioned in my earlier post.