Of course it would need to be implemented after SMTs were up and running and giving Devs and business owners a bit of time to add value.
Yes, it would need some time, but keep in mind this post is merely to start a productive conversation. Because, as you said yourself, a vision is important.
However, why would we need to change curation rewards and add a flag pool if we went this route?
Those would be added into SMTs, where each SMT could decide how high author/curation-rewards should be and if a downvote-pool should be included. (so the downvote-pool would have to be coded anyway) E.g. Steemit Inc could decide to use their own SMT for Steemit.com.
Why would serious business owners and devs come and build on a blockchain that can't even decide what it is, what we do and how to approach it.
Because every serious business owner and dev should know that things change over time. New features are being introduced and old/deprecated ones removed. And Steem has clearly changed quite a lot over time - at least once SMTs are there.
Also, everybody can launch their own project, but steering it through tough times, that's where the real value is shown and where trust is built.
I agree with several things you said, the flag pool and curation rewards belong on an SMT not Steem.
Yes, things change over time, but why customize STEEM to work better for curation rewards and fighting abuse if the ultimate goal is to use it as a base token?
Things change over time, but messing with inflation and curation rewards has the potential to impact business models, profits, etc. That's a pretty big risk to take on a blockchain that doesn't seem to know if it wants to be a base currency to trade SMTs on or rework the incentives to be a content discovery platform.
Both discussions are important (changing the economic model & the base-currency vision). The idea of reducing inflation drastically is actually not new, because we gotta ask ourselves: Is the currency distribution working - and if not, is it necessary? Or is it more evil than it does good.
Now, the good thing is, once SMTs are there, we can test out different economic models. There could be a curation-SMT, where curation is at 100% and no author rewards. And maybe an SMT, which works 50/50 or 25/75, as it is right now.
All true, and worthwhile experiments, but where do the SMTs get the rewards so that they can change the reward distribution when, in your model, there will be no author/curator rewards?