Am I the only person surprised by the conclusion that far enough out, with reasonable assumptions, supply is independent of starting amount and only a function of emission and average loss rate? Not arguing with the math, just surprised. Of course, I'm sure people have debated the assumption that lambda is constant. I could imagine an argument for it shrinking over time, maybe more careful custodians become the norm, who knows? Also I guess we're ignoring the related topic of mining environment/incentives between scenarios 1 and 2. Option 2 feels nicer to me, but could be my bias for upholding the spirit of original code.