Parallel sounds infinitely less complicated so +1 from me on that. The time I think needs to be considered carefully as you rightly state. Too easy to end up in some kind of CypherFUNK gold rush, and I'm not certain of the effect that could have mid/long term.

Given that it's going to be such a simpler system to manage on top of XDP or whatever token issuance tool is chosen, I think it's important that thoughts into the applications which will leverage the token are put in pretty soon - the development of these now becomes far more a core process than the management of the token itself. They *are* the token's success.