The logic behind what I'm saying is why spend time and resources on something that nobody wants? Or that a tiny minority want. Getting feedback is free and fast. Maybe the feedback would be overwhelmingly in favor, but without that your just adding features at random.
Features need to be developed around user demand, not just to see if they'll work IMHO.
If only a tiny minority want it then witnesses won't upgrade (and those who do will likely get voted out) and it won't go live.
There has already been some informal consultation (and both recently and in the past some on-chain discussion) with witnesses and large stakeholders which suggest it has a legitimate chance to be adopted, though I also wouldn't rule out that it won't.
With that in mind, if we say it has a (being generous) 30% chance that it wouldn't get approved, are there other features that there is widespread consensus on that have a 90-95% chance of getting approved?
In a nutshell: Will this take time away from Communities or SMT's because that's the stuff EVERYONE wants. Why not just push full Steem ahead on that and after that's out we tweak all this stuff.
Communities are not even a blockchain feature. They are planned to be implemented in hivemind as far as I know, which is a layer on top of the blockchain.
SMTs development isn't finished and I don't know when it is finished, although supposedly that is the next thing to be worked on.
As @baah noted, this particular issue (downvotes) isn't a major coding task either way, but there is a lot of support for some ways of improving the function of the Steem economy.
As far as trying to put percentages on specific features I don't really know. I think it is sufficient that developers don't waste their time on things that have little chance (and that has happened in the past) but I don't really see that here.
Just limit the amount of funds that can be extracted from rewards. Sound investments are invariably based on increasing the value of the investment vehicle, and rewarded by capital gains. Investors have been so encouraged since prehistory, and this is the basic mechanism which has created our extant markets.
That will end bidbots, self-voting, and such without doubt. Ending extracting rewards by manipulating curation will allow curation to actually be based on content quality as judged by individuals, rather than parasitized by profiteers.
Anything else will continue the downward spiral.