I've done a bit of reading on the links you posted above, @steemitblog's and @vandeberg's latest posts, and others, since we spoke last.

It seems your persistent efforts have born fruit, as your proposal is essentially what Steemit is bringing to the table.

However, like many tweaks to the rewards previously undertaken, I note that these are just tweaks that do not address the underlying causes of the basic problem. The problem is that the manipulation of curation for financial gain is the means of attaining ROI on stake.
This then impacts curation and degrades it's intended purpose as a means of encouraging creation of high quality content, which is basically advertisement for Steem and a means of attracting capital gains inducing investors, by introducing a stronger motivation to extract immediate ROI from rewards via financial manipulation.

These tweaks don't change that basic problem. Extracting rewards as ROI is contrary to investment purposes, as it reduces upwards price pressure and thus potential capital gains. Were this not afflicting Steem to date, I am confident we'd see considerably more capital gains, and not this gradual decline in market cap I've observed since my introduction to Steem.

You may not recall how I proposed to counter this basic issue, so I'll briefly recap here: create a mechanism using the SPS (SteemDAO) system to provide a dividend stream from delegating and funding basic development of Steem and it's ecosystem for substantial stakeholders. Presently the extraction of rewards via self-voting and bidbot delegations puts downwards pressure on Steem price, directly suppressing capital gains. This dividend stream reverses that effect on capital gains.

Secondly, remove the incentive to extract rewards by financial manipulation altogether by limiting rewards to a multiple of the median. Leaving no cap on rewards potentiates abuse by substantial stakeholders, and this demonstrably has occurred. Capping maximum rewards to a reasonable multiple of the median eliminates this by making it impossible to scale extractive financial manipulation nominal to interest substantial stakeholders. I've proposed a model suggested by Huey Long: 3% minimum and 300% maximum of median rewards as lower and upper limits, since this leaves two orders of magnitude for differentially rewarding content per it's quality as judged by curators, yet also makes it worthwhile to even new users to post. Many new users receive literally nothing in rewards for their early posts (since they haven't yet created a network) and quickly get discouraged and leave. Many, many mechanisms seek to address this problem, from a significant number of established accounts, and this would aid all their efforts in attracting and retaining new users.

These changes directly address the fundamental problem, rather than merely tweak it. Please give it thought.

Thanks!