I think dismissing it because the current implementations are still startlingly awful after several years, with really obvious feasibility problems for their po

I think dismissing it because the current implementations are still startlingly awful after several years, with really obvious feasibility problems for their posited use cases, is fair enough really. At this point the burden of proof really is pretty much entirely on the blockchain advocates. Bring a technically feasible use case to the table.