Those of us who see steemit as a social web site that happens to use steem to pay out rewards for content might be tempted to focus exclusively on the importance of rewarding high-quality authors, and that is important, but there's a different perspective to be had. This graph surprised me.
Source: SteemData - Blockchain Operations Distribution
This snapshot was from a couple days ago (when I saved this draft), but most days look similar. In terms of transaction volume, 92% of steem blockchain activity is about voting and rewarding voters.
The authors have the biggest "soap-boxes," so here at steemit we hear a lot about what makes them happy and sad, but as a community, we should not lose sight of all the different stakeholders. We should not let the squeaky wheel be the only one to get the grease. Even though voters don't have as much of a voice as authors, their prominent role as stakeholders should be remembered.
In terms of value, after the reverse auction for early voting, the author rewards are reportedly around 7 times the total value of curation rewards, but - conversely - in terms of distribution, the curation rewards are distributed to about 35 times as many recipients (not necessarily unique). Simply put: Comparing the two pools, a far smaller curation pool is being distributed in a much larger number of disbursements. It is clear from this data that curation rewards are far more efficient at creating steem transactions than author rewards (by a factor of about 245).
If your goal is decentralization, then it seems that curation rewards may be doing a better job than author rewards of meeting that goal - which should not surprise since there are more voters than there are authors. If your goal is to bootstrap the steem block chain, the transaction volume shows that curation rewards are accomplishing more than author rewards. If your goal is to incentivize holding steem, curation rewards incentivize voters infinitely more than author rewards incentivize authors - because aside from power-down constraints, the incentive for authors to hold steem is zero.
Personally, I don't see a need to adjust the curation reward pool at all at this point. I think it is far too early to judge how things will work in the long term, especially while we're still in the power-down window after Dec. 6. But if we're going to talk about adjusting it, the chart above seems to suggest that increasing the curation reward pool, not decreasing it, is more likely to create buying demand that will raise steem's value (and thereby raise payouts to both authors and voters).