So what's the solution? 

It has to come at the coding level, no?

Do you think we have enough visibility yet? To me, we could do more. This could be a regular topic of discussion with actual data. Weekly reports (maybe even daily reports?) and ask for engagement from the voters. I too often see people complaining about the author without calling out the actual voters who are doing the distribution from the rewards pool. Why don't we focus on the voters more? Why don't we have reports that don't assume blame, but instead try to explain? There may be valid reasons why some people decide to add others to voting bots. If we don't know those reasons, we can think it's a scammy voting ring. If we do, maybe we can reset our expectations on how (and why) people use their own Steem Power as they choose.

Many of the examples I've seen of people caring about the rewards pool distribution are (IMO) immature comparisons to "rape" instead of rational, non-judgemental discussions trying to understand _why_ certain people vote the way they do.

People have reasons. Our ignorance of those reasons doesn't justify our moral judgements. If, on the other hand, we know those reasons and can clearly articulate them in ways the community at large can fully agree (or disagree) with, then it's possible support for those people will be removed or added. If there's still disagreement about the value behind the reasons for a vote, then it's not an issue we consensus on, so even more discussion should take place before we claim the community has figured this out and our solutions to it have failed.

A great example is that my girlfriend is on Steemit and I know that if I don't upvote and resteem what she writes she'll get mad. Fortunately she puts a lot of work into what she writes so I don't feel bad about it, but it's a microcosm of what you are describing.

Why not accomplish both?
The action to upvote is levied by [your] account's reputation score, an index of your blockchain voting history (collective upvotes, comments, and resteems). Should justify reason plenty, however, there are larger forces at work that reason escapes the rhythm of logic. 

I am unaware of sorting algorithms for the feed, similar to Facebook, in action here on Steemit, but if there is, it's much more flexible than the big Blue page. Steemit could employ automated bot services and controller/curators to divvy the flow of traffic, content, and category. SMT seems to offer at least partial complex to find a solution to financial exchange for content type, perhaps Proof-of-Brain concept may light the way to truly thoughtful, honest, information to earn author's weight in value.

<center>
<br>
### *Time is Art. In Life, all we have is Time. Why not make Art all the Time?*
</center>
- - -
 <center>
![ASC_EN_Original.png](https://steemitimages.com/DQmaSE9PqEMnYdPmVUuU4pw5mqMUoGTzDdBd7VV8xdpG2m8/ASC_EN_Original.png)
I am a proud @earthnation Steemit Guild Community member!
Thanks to their loyal support, I am able to create a passive income stream that funds my creative output and delivers this original content directly to YOU.
</center>

There's also this useful quote to consider:

"The only thing necessary for the triumph of evil is for good men to do nothing."

I think it takes both positive reinforcement and clarity on what is accepted and what is not based on action which includes downvotes.