People are always like “it doesn’t do anything we couldn’t do with SQL,” as if riding a horse isn’t an improvement in transportation over walking. Things don’t have to be impossible to accomplish in any other way in order to be marginally more useful or efficient depending on the goal. Public ledgers are indeed useful. Blockchain is one technology among a small handful that might be appropriate for your project depending on trust dynamics it demands. Consensus protocols are also useful.

One example, right now our global food supply’s movement and distribution is based largely on market dynamics. Say we want to focus on distribution based on need instead. A blockchain based ledger could allow a fred to ‘commit’ a few bushels of carrots which george ‘commits’ to transporting to mike, who in turn has committed to do do a supply run to Uruguay with his barge. Could they have done this in excel? Probably. Would it be more organized on blockchain? Yes. Would a regular database with a lot of contributors that is carefully designed to keep out bad actors work too? Yes, sure.