That strongly suggests that most "legacy" Steem behavior will not be supported going forward. There are very few existing features of Steem that forward those goals. A development roadmap with that as the endpoint ought to start by removing the Steem rewards system wholesale, as an extremely expensive feature that none of the long-term goals actually require.
It also implies that everything anyone's building for the "first-layer" Steem right now will become obsolete through the SMT development process.
Would definitely like to know where I'm wrong about this, because if that's your mission statement I probably shouldn't continue working on projects here.
“Interactive currencies” includes STEEM and SBD
So the blockchain of your vision is going to run very similar content-voting-and-rewards systems on multiple levels of abstraction? That seems anathema to the efficiency-and-scaling doctrine.
Once you've built new voting systems and dropped dApps into their SMT walled gardens, I think it's going to be way too tempting to drop first-layer content voting and rewards and just have the core blockchain support content storage and SMT operation, while all the rest is done per-dApp on the SMT layer, in the name of chain efficiency. And it's probably the right thing to do from a system design standpoint.
It kind of calls into question why starting from Steem is a good idea, though.