Yes, some good feedback there. On the first point: I've always thought that this kind of thing specifically addresses the most stringent privacy requirement (or, tries to). So if you consider things like DOS on LN or Tor or other similar networks, LN micropayments can def solve it, but they can't achieve the same anon properties as anonymous usage tokens of this form. In practice, does that matter, given that LN has a decent chunk of privacy built in (if not perfect)? I think only for certain scenarios would you say, no, that's not good enough (there's also a bit of a bootstrap problem to defend LN with LN).

Ecash is faster but suffers from the trust and network effect problem, which are intertwined. Which mint? Crossing mints means more complexity and trust issues maybe? But again, you can imagine niches where it might be ideal.

I'm not going to comment so much on the web of trust kind of ideas that you mention you're working on, but i can imagine them having big advantages w.r.t. user not having to do anything, compared to some commitment like a payment or a token of ownership.

Re: proof of rich, ultimately I *suspect* you're right that it's more useful for higher (not necessarily rich but at least 'larger amounts') size money proofs/commitments than for small or micro- ones, but one never knows how a tech finds its use case until you see it in action, I guess.