That's the basic idea. The question is whether it's cheaper to lease the additional voting power needed every time, or cheaper to lease a little extra up front to provide upvoting rewards. With our 'above average' curation rewards the extra expense up front is covered.

Realistically we should make upvoting rewards a parameter too (set at author rewards + curation rewards, but using our curation % instead of assumed 25%). In this case that would be 75% + 37.5% = 112.5% as a sustainable upvoting reward multiplier. This is something I will investigate further.

Making the different reward aspect paramaters that we can manage based on the performance of different areas is easier for us to maintain sustainability, but more complicated to explain to members.

Instead of 'the reward is always this' in the FAQ, it would be 'the reward is based on this complicated formula and constantly changes'. It's better to overpay a little and build a reliable base of upvoting members than to overmanage.