"The issue is: current curation/reward system simply does not work. "
I think this is a big part of the problem. The curation system should encourage upvoting others more. However, from my own observations, it rather strongly encourages upvoting yourself or creating bot nets. Maybe adjustment to the 30 minute rule is necessary, or maybe curation simply needs to have its overall coefficient reduced.
Either way, curation is hard to abuse and I think it should be rewarded more.
Curation should be an incentive to find and upvote quality content but as it is people try to maximize it by following and uptvoting authors (not content) that they expect to get more rewards. Besides that, the best efficiency on upvoting power is only achieved by bots (or an extremly disciplined human) and this also benefits bot curation which hardly ever is content driven.
For it's purpose, curation fails to achieve what is necessary for steemit to grow, that is making people see it as a usefull and fair plataform, where you can find good content easly and get rewarded by helping others to do so.
Great point about the Bots. They are fundamentally experts at curation, since they can work 24/7 and keep exact timers, plus they can analyze data and pick based on expected reward, not on content quality. Bot curation probably makes manual human curation almost totally pointless.
Bots are experts in maximizing curation rewards by time adjust and most probably(*) work on author/elapsed time selection plus some social proof of quality - but not content. There are a few exceptions, like gentlebot, but not enough in my opinion.
(*)this is a huge guess based in what I see, since I can't check the algoritms
I belive without real content driven curation steemit won't succeed. I belive it because without it new users won't see the plataform as fair neither find good content easly. A lot of posts about this subject are comming to the surface and I can't see a solution in the near future, so that's why I am less confident about the future of steemit.
Hopefully I am wrong and/or the community will find a way to fix it.
The truth is I think Steemit has a lot of real content that's actually pretty good, but it is buried under a ton of garbage and another layer of misleading-highly rewarded OK-but-not-excellent content.
Gentlebot is great, I've noticed he seems very effective. I love when he hits me up.