As a computer scientist, I love it. It’s a fundamental breakthrough: the first open participation distributed consensus algorithm ever. That’s a big deal.
As an engineer, I can’t responsibly recommend it to anyone for anything. It’s slow, unreliable, immature, hard to use, and functionality impoverished. It’s a database that hates you.
Truly heartbreaking.
But blockchain is not (just) about open participation. It's about trustlessness ("well actually" there are "permissioned blockchains" but that's just a glorified/terrible variant of Git pretty much). And society is built on trust. As Nicholas Weaver puts it, blockchains (trustlessness) is good for one thing only: censorship resistance, i.e. evading the law. It's just the only option for thing like ransomware. For everything else, systems with trust are many many orders of magnitude more efficient.
And I'd say systems without trustlessness can be "open participation". Dename was a pretty cool idea.