Having the community to find good posts is an innovative way in my opinion of distributing the rewards on the blockchain. Having curated now for a week I wanted to give my little review of this feature.
First of all, it's really great to be able to give some big votes to people. If you have been a minnow for most of your time on steem you know that your votes don't really matter too much. So this gives you the chance of really making (a little) difference. Rewarding people is fun, and I would argue that this is one of the things that makes steem worthwhile in the long run.
Although it is fun, it is also really work. This is because you obviously have to browse through newly created content and check for every post if it should receive more rewards. This obviously takes a bit of time if you want to do it seriously: you can't just curate random posts. So I think some people might realize that having more (voting) power and distributing it to quality posts is something that they don't want to do. If you are not willing to spend your time on this, it won't work out. But that's totally okay because it's not something that everyone should necessarily do!
When you decide on which posts should get a curation, you quickly realize that there is a dilemma. There is often good content out there from authors who usually just receive a very small amount of payout. Voting for these posts will probably mean that the efficiency (actual curation - standard curation) will be much less than for voting on posts that generally receive a larger payout. Therefore, it is very tempting to vote for the big ones out there to maximize the efficiency of your votes, but this will in turn distribute the rewards to the authors who already receive the most. Minnows will again be left out of the rewards pool...
The danger is that this can lead to a very one sided way of browsing through /created: look at the authors reputation (=> if >60 look at their profile), then check their post history for rewards (=> if >5 Steem cast vote for current post). This obviously isn't the kind of behavior that should be done when the general idea is to spread rewards more evenly to the community.
So how to deal with this problem? My rule is that the posts should always have some subjective value to them. Simply voting for posts that will probably receive a high payout is not improving rewards distribution (which I think is one of the goals to make steemit better). So one actually has to spend at least some time for looking at the post's content. If it's subjectively good content then simply cast the vote. It's as easy as that. Casting a vote will increase the likelihood of other people seeing the post and this in turn might bring up the rewards even higher. However, to balance the potential losses I would suggest to still vote for posts that you know will receive a larger payout (but also have good content!).
So why are we doing this? Well, to be honest the rewards are modest coming from casting votes for 
These are my 2 cents on this feature. I'd be interested to hear from others who have joined this experiment :)
