Wanted to share an update here: I’ve chatted with some more folks and – under guidance of the maxim “PMF above all else” – I think it might be good to adopt a “two-phase” approach here to deployment in order to allow frontend operators to test what fees make most sense since they have the most direct relationships with users.
This would mean that instead of upgrading the protocol right now, we add support in the Namada SDK for “customizable withdrawal fees”, so that frontend operators can add a second output when the user shields or unshields which goes to the frontend operator (and they can pick what fee to set and what schemes to experiment with / harmonize those choices in a way that their users are happy with). Then we can collect some real user data from decentralized frontend fee experimentation, get clear evidence of what scheme(s) make most sense, and later upgrade the protocol to split the fee.
Thoughts?