Sorry for the late reply. Spent the past weeks in bed with a flu. So limited myself the last few days to doing only the most essential stuff (downvoting shit posts). And I thought your questions deserved a more thoughtful reply than what my feverish brain could put together.
First of all, I was in no way trying to make a case in favour of ocdb as something positive. It is still inferior to having the same stake used by individual curators to curate content. For the reasons you’ve stated, which I agree with 100%, but also because less people participating in the intended curation game leads directly to less engagement. Having people spend time finding content to vote on it makes them more likely to leave comments, and thus more life to the platform. It’s just that ocdb has a few less negatives than other bidbots, since it doesn’t contribute to the problem of promoting garbage to Trending the way that others do. I just thought that was worth pointing out to not lump them into the same bag.

In any case, I believe serious economic change was needed, which is why our witness (@steempress that I co-run with @howo) was in favour of the EIP back when Cervantes did the poll. (I don’t think we were as close to the top 20 back then, so perhaps our opinion wasn’t noticed by you). We’ve also been vocally against bidbots and in favour of manual curation and community use of downvotes. That’s why we recently launched the downvote management tool publicly despite knowing it could result in backlash against us. It’s also why for a year now we’ve supported 24 different manual curation projects and communities by trailing their vote on content published through our plugin. This to 1. Help make curation more rewarding and thus competitive with content-agnostic alternatives (self-voting/bidbots), and 2. Because we want to support the idea of proof of brain by letting a wide variety of curation projects looking at different topics contribute to our own curation.
I was hoping to have a chat with you about it, since we’re about to launch a new referral program and curation dashboard where people can support active curators. I would love to know which subjects you would be most interested in seeing us work to onboard, and if you would be interested in curating content from those areas should we attract quality creators.

So back to the questions about the EIP:
My main beef with the fork is mostly the fact that all the changes got lumped in together at the same time. I know they are intended to be complementary, and have read your post, but imo this makes it hard to isolate which of the actual changes have made positive or negative effects on curation, behaviours, and other KPIs. Perhaps it’s just the scientist in me who prefers a cleaner experiment! I also mentioned that I am concerned about the effect of the new curve on smaller users, communities and the curation of good comments. As you’ve said, there will be some trade-offs on the way to a system more functional, so this is where I would question if the trade-offs could outweigh the positives of that part of the EIP change.
At the end of the day I am more focused on the behavioural economics than the macro at this point. I want rules and incentives that are aligned with a growth strategy to improve user acquisition and onboarding above anything else. That’s why I believe we should be careful to make changes that have trade-offs which negatively affect the grassroot communities. Growth always happens at the margin, so even if the EIP reflects the best possible end-solution, one has to be careful with the short term implications if they result in user frustration or reduced motivation to stay at a time when users are most desperately needed.
With that in mind, if I were to look at the benefit / cost aspect of the different parts of the EIP that make the most combined sense right now. I may have suggested first going for free downvotes and improved curation rewards. Imo those are the two that contribute the most to provide the benefits that I know both of us wants (punishing content agnostic bidboting and circle jerking while incentivising good content discovery). I just think that most of the benefits we’ve already seen over the past week could have been accomplished by these changes alone. So while they are responsible for the benefits, I see the curve-change as the cause for most of the negative side effects without contributing as much to solving the problems as the downvotes and CR do (although this could only be short term).

I am also vary of how while solving one problem, we may create new problems when it comes to having people “vote honestly” to reflect their subjective judgement of what is good. With linear, it seemed more people were happy to just vote on comments they found nice, or small posts by people who they appreciated having on Steem. While arguably not contributing much to content discovery, I still see this behaviour as valuable in retaining users and making Steem a place where people find conversations and engagement more valuable to have. So if these votes starts turning into users instead auto voting the same popular creators, who usually get high rewards, at the 5 minute mark to get more curation rewards, then it will be a very costly considerable trade-off imo (like this example here https://steemit.com/newsteemit/@fishyculture/my-list-of-whales-to-curate). I even find myself doing it (although only on creators I find worth something). I’m not sure if you’re fine with that, assuming the content they auto-vote at 5 minutes is seen as quality by the community and not downvoted? At least in my mind, I would rather have many users just voting what they believe brought value to them. As I think this is good for user retention. When people try to get ahead of the pack on the popular Steemians, it results in very few getting any rewards and getting seen. So I would just like to see a system where people are more focused on 1. Curating whatever they liked for no other reasons than that they liked it (honest curation). And 2. That the way people vote helps improve the positive experience of more people on Steem, improves user retention, and thus allows the platform to grow.
So I guess my preference would have been to introduce the free downvotes and increased curation rewards in this fork, and considered having different reward curves with SMTs to observe what makes the most sense.
In any case, this is not arguing against your model when it comes to how we can improve content discovery, but it takes a few other factors that I see as important into consideration. At the end of the day I believe the only way to tell what works is to give it a whole-hearted attempt, which I intend to do and am doing.

And goddamnit, now I can't get that comparison out of my mind..