From your first post, I've been thinking about one particular excerpt. This is almost an argument in favor of the bidbots:

The reward from this single vote will eclipse the equivalent that you'd get through self-voting so as far as maximising your returns (your ROI), it's in your interests to submit 1 post per day, every day.

So if you consider the low-appeal content as a form of "pollution", then the argument can be made that they've reduced pollution by 90%, since there could, otherwise, be 10 times as many low-appeal posts.

In today's post, this is a good point:

If a proposal was voted to the top of the list, surpassing the "Return Proposal", there's a limit to how many SBD can be withdrawn which will allow time for somebody to react.

Also, to control the SPS, they would have to have more SP than Steemit + all other non-proxied voters. It's not just a matter of out-voting Steemit, so getting unilateral control of the SPS is probably still far away.

In addition to the potential improvements above, we also have to give upvu credit for innovating with their @steem-punks. To my knowledge, not even Steemit recognized the blockchain's 6^th anniversary last week, but steem-punks did.

Basically, services like UPVU and other bidbots turn Steem into a proof-of-stake rewards system, so maybe we need to turn the question on its head, and ask, "How can we get better than proof of stake at delivering profits to high-appeal content producers?"

In the end, less pollution would be better, and 0 pollution would be best. I'm not gonna turn around and start delegating or proxying to them any time soon, but this has been a good exercise. I'm starting to see that things aren't as clear-cut as they might seem, superficially. At any rate, I'm less alarmed now than I was when I read your first post.

Now, excuse me while I go learn about the Coase theorem ;-).