> As I understand it you just need a hard enough computation to prevent the 51% problem. that's not the case, in fact hardness is barely a requirement provided

> As I understand it you just need a hard enough computation to prevent the 51% problem. that's not the case, in fact hardness is barely a requirement provided you can scale up the number of sub-tasks as much as you need When the work is provably wasteful, the model simplifies a great deal. Usefulness of the work complicates the dynamics and makes many scenarios plausible that otherwise wouldn't. The game theory then doesn't follow that you'd choose a currency with extra elements of instability. I don't think there's enough evidence to make the hard requirement that it HAS to be provably wasteful. It's just that you introduce a bunch on extra, usually intractable, problems when the work is useful.