> Social consensus, however, doesn't affect the standard 'malicious' double-spend cheat. You're not going to get social consensus forming around an obvious 'criminal' behavior.

You haven’t refuted my point with any evidence. I’ve shown how one could spend a digital token twice by taking advantage of weak social consensus (see my Luke DashJr example). This attack bypasses Nakamoto consensus. You don’t need any hashrate or material wealth to pull this off, you just need social capital... or a wrench https://xkcd.com/538/