I think you make a number of excellent points, particularly, “relationship are always corrupted or distorted when a third party manipulates for that third party’s benefit.” So what if the social media platform was owned by the users? I see two immediate consequences: 1) Some users might try to exploit other users just as they do in the real world but there would be hard to protect against because of no central source of responsibility; 2) Without strong corporate backing, the platform would languish because no one would take responsibility for it.
Therefore the owners of the platform need to make enough profit to maintain the usability of the platform. They could follow a subscription model and this would be most likely to be successful if it could appeal to particular interest groups who would be willing to pay for quality interactions that are not exploitative with others who shared the same interests. Many would likely be willing to pay just to get rid of the trolls if there could be a way to escape them.
Just like we try to form a circle of quality around ourselves in the real world, could we find way to do the same online? If the selection process was driven by human decisions rather than algorithms, the quality could be constantly enhanced because it would grow through experience and interaction rather than keywords. The real problem is that we have been very effectively trained to use all online interactions as resources to be exploited for a number of motivations, not all of which are monetary.
It seems like its the same problem we face in the offline world. What are the social tools we use to build up our humanity, to govern ourselves so that we build the solidarity we inherently long for.