Yes. It assumes the asymptotic end condition, because to me that is the only reasonable way to judge such policies. Cases of "nothing was done because bounties

Yes. It assumes the asymptotic end condition, because to me that is the only reasonable way to judge such policies. Cases of "nothing was done because bounties weren't interesting", or "we had a feature bounty program and it done awesome!" are, for the purpose of my analysis and thus opinion formation, only more or less happy exceptions (hypothetical or real).