Metacuration is the act of curating curators; in the case of Steem, this means building the stake of curation accounts so they can vote on a broader range of content and distribute rewards among a larger group of people than I have time and attention to do myself. Let's take 
I vote on the PIFC account, as well as a bunch of other things. The PIFC account votes on its curators, who in turn vote on and interact with new users. This allows my stake, eventually, to reward those new users by making each account in the chain larger. The process is glacial, in that it's very slow but it eventually makes a large impact. This is what I've worked to build on Steem, and this is why 50% curation rewards are going to drive me away when it happens.
What happens to centralized stake moving down the line under increased curation rewards? Here's a chart of a four-level metacuration and account stake under 75/25, starting fresh - a million-SP account wants to spread the wealth to other curation programs and eventually end users starting from scratch.
| Year | Founder | Curator 1 | Curator 2 | Curator 3 |
|---|---|---|---|---|
| 1 | 1045615 | 142724 | 9925 | 468 |
| 2 | 1094268 | 298733 | 41166 | 3817 |
| 3 | 1145184 | 468950 | 96629 | 13358 |
| 4 | 1198470 | 654362 | 179496 | 32981 |
| 5 | 1254235 | 856013 | 293233 | 67222 |
As you can see, the growth at the bottom is slow, but recursive. As each curation account gathers stake, it can go forward and pass on the value to others.
Here's the same chart under 50/50:
| Year | Founder | Curator 1 | Curator 2 | Curator 3 |
|---|---|---|---|---|
| 1 | 1093269 | 99484 | 4616 | 146 |
| 2 | 1197328 | 217910 | 20024 | 1239 |
| 3 | 1311292 | 357979 | 49181 | 4533 |
| 4 | 1436103 | 522737 | 95599 | 11711 |
| 5 | 1572794 | 715617 | 163433 | 24979 |
Things have changed around. The starter account now has 320,000 more SP, man, I'm sure that's great for them. In five years they've increased their holdings by 25.4% over the 25%-curation system.
In order to pay for it, everyone else has become less effective. The first-level curators are the best off, they've only been cut by 16.4%. In Year 3 they managed to get enough SP that they're reaping some of the benefits of 50% too, so they're actually starting to catch up; they'll match where their 25% counterparts were somewhere during year five, and actually surpass them around the end of year 8.
The second-level curators got a little bit hosed, they're down 44.27%. Almost half of their effectiveness has been stripped, but they're starting to get some benefit from curation here in year 5. They'll be where 25% Curator 2 was midway through year six.
The third-level curators were pretty much destroyed. They've lost 62.84%, and moved from an orca to a middling dolphin. They'll catch up to their 25% counterpart toward the end of year 7.
It gets worse if we add more levels. Level four drops by 75.22%. Level five drops by 83.48%.
It also gets worse if we add more years. For every year out, it takes longer for the accounts under 50% to catch up to where they would have been under 25%. You can see this in the chart - in year 2 Curator 3 was less than one year behind his 25% counterpart, by year five he's almost three years behind.
This is the long-term stake distribution change being promoted on Steem through the 50/50 rewards proposal, simply and obviously. Do you wonder now why I think it's terrible? If you're a top dog and all you care about is your own ROI, you'll get a nice little benefit. If you're not a top dog, or you care at all about your effectiveness, you're dead meat.
(By the way, these numbers include recursion for everyone reinvesting liquid rewards and voting with their curation SP. Guess what happens if anyone cashes out.)
Assumptions:
- current amount of system inactivity
- curation rewards at par (25% of vote at 25, 50% of vote at 50)
- all accounts use all votes
- all rewards used to power up
- either 100% powerup or SBD stays at $1
- I compounded weekly rather than trying to guess when posts happen.
Based on further thought from this I've developed an alternative proposal which would allow natural discovery of optimal curation levels, and I believe should work to the benefit of all stakeholders.