The ideas that come to mind when I read this is:

If you look at steem there is scopes.

You can look at posts as the smallest scope. Posts are what what you want on steem when a user is active and each post is different from the next, unless you evaluate them by bigger scopes.

Then there is accounts, accounts are generally a bigger scope of a posts activity, your bad accounts will be very likely to remain problematic throughout more than one post, rather than creating one bad post they would typically create many, same goes for good accounts, they tend to stay good.

Then there is groups, groups of accounts indicate that we can possibly draw higher conclusions than the stream of posts that an account producers, there is the stream of accounts that groups produce.

So while groups are difficult to always determine we have accounts to go by, so beyond rewarding posts with votes, one could look at producing phenomenon that rewards accounts that look promising.

Beyond that we can say that if you become a member of certain groups, it might be fair to receive the attention that other group members have. So we could look at maybe software that allows groups to host movements that fund member participation, and so set up guilds where exceptional new members can, instead of receiving the curie trail vote, get bonus delegation and capital incentive on a more permanent basis alongside invitations to group memberships for more permanent boost for good members, which widens the gap between good users and abusers.

I think there is a lot we can do to create guilds and groups that boost good members, decentralizing the moderation effort so that groups of relevant interest can self moderate and use their group power to create footholds for certain communities on steem.

In some arenas up to date information is a major thing, steem has a way of featuring new stuff that if we simply establish communities that reward information that is totally new, we'll already give steem a huge boost as a network.