I could create "Distributed Render Service" application - you register your machine with my service. My service runs on a pretty standard set up on AWS. I keep the details in a sql database. There is a client you download to your machine which basically figures when the GPU isn't in use and makes a call to my service for a "job". The client pulls down the relevant data, completes the job on your GPU and sends the data back. My service credits your account. On the flip side I have a place for people who need said rendering done. They upload "jobs". My central service figures which jobs to send where.
Where does a blockchain make it cheaper, more compatible or improve service? I understand that you state that a blockchain isn't the only possible means - but the key difference in question is central, trusted control v not. My service has a central, trusted control. Blockchains are created to avoid a central, trusted control. So even if we avoid blockchain - what would be the scenario where you can concretely avoid central, trusted control and it would be beneficial on the dimensions stated?
Fwiw, I believe you are making these statements in good faith, and they aren't argumentative for their own sake. Like the original poster, I just struggle to see the value in 99% of the situations people propose involving blockchains (or avoiding central/trusted control).
