On Steem, the reward pool receives the largest part of the STEEM inflation. Even after the economic changes, 65% of the inflation still goes to the reward pool.
That's why we have these fights over rewards, because they are the biggest slice of the pie, and the easiest to abuse.
So, yes, I believe we shouldn't have a reward pool for STEEM anymore. And if you read the comments on the lastest steemitblog post, you'll understand this is being considered at some point.
The big problem is when you do it. Not if you do it.
It seems like a necessary step in the future.
Yes, as I mentioned in a previous post, removing the reward pool for STEEM won't stop abuse and wars from moving to communities and tribes.
But hopefully, in smaller, more targeted groups they will be easier to manage. Plus there are and will be ways to deny bad actors access to rewards, which will remove the incentive to have the abusive behavior, for profit gains.
It will also lower the inflation on STEEM, which in turn might reduce the downward pressure.
Let's return to the when part.
When have tribes been introduced?
Palnet is the oldest but the most established ones, all have at least a few months, I believe.
They are still in their infancy trying to define themselves and grow, in my opinion, even the best of them.
Here's a metric I would use for a tribe/community to make it on the long term, if they don't already have a strong community behind interested to grow its economy: when people who have no relation to Steem and crypto start to find their website (interface), see the content and all the other benefits, and feel they want to join, and they do. On their own. Organically.
Then such a tribe / community has authority. Its voice is heard, and listened to, and with that its economy flourishes.
Also SMTs will go through STEEM to exchange to anything else. For a SMT and its community to make it on its own (and for the benefit of STEEM, in most cases), they need direct pairs on exchanges, not through STEEM.
HUNT token is an example of token which is listed directly on a few exchanges, and they will have a STH token as SMT.
Before we have such communities / tribes which can safely stand on their feet, I believe removing the reward pool is risky. Because we put our trust in something that hasn't been proven to work yet as a good replacement for the main reward pool.
There is also the risk of an increased downward pressure, which plays in favor of removing the reward pool. If earnings and profits from the newly created SMTs are quickly turned into STEEM and sold on the market, that adds to the existing selling pressure.
I'm not sure what the fees for creating and / or managing the SMT will be, but they should create some buying pressure to compensate. That, plus the needed RCs for the community accounts, since creating a Hive Community will only cost 3 STEEM (!).
If the potential downward pressure via the newly created SMTs is not too much to handle, I don't think the STEEM reward pool should be removed in 2020 for sure, to allow tribes and especially communities the time to grow and establish themselves, without putting all the pressure on them.
A few big websites adopting SMTs and having their own direct pairs for their SMTs on important exchanges could be another valid reason to speed things up on removing the reward pool for STEEM, though.