> by giving up on the “fixed known supply” model in favour of the “this is a product that is sold in perpetuity to consumers who find it useful”
The philosophical problem with this approach is: sold by whom? These projects are generally designed with the philosophy that the central creator is only there to bootstrap, and the project should be eventually autonomous. So the sale opportunity has to disappear eventually.
Now, you could say that the token could be sold by a DAO, but then that DAO will have to have a control mechanism for directing funds from selling coins, and the value of controlling that control mechanism will be equal to the expected future discounted token sale revenue, and it’s difficult to avoid that control from being tradeable in some way. Hence, what you end up with is the volatile coin / stable coin model: https://github.com/rmsams/stablecoins . Now, should we have 500 instances of volatile+stable coin pairs for 500 applications? That depends on what you think the purpose of these tokens is, and is a more complicated question.