I really don't think this is going to work. This is an idea whose time hasn't come.
I tend to agree. I don't think the technology is here yet. Literary types will be highly critical of this idea. An editorial built this way will either have no voice or contain such a cacophonous jumble of voices as to be considered psychotic.
A wiki is the wrong tool for a collaborative editorial. It crudely combines too many discrete steps to be manageable, especially at scale.
What's needed is:
1) collaborative position finding and group formation. Active participants need to be able to express approval or disapproval on a statement by statement basis. Like-minded people must form groups and work together to find the most effective way to express their sentiments. In the wiki format, it's all one global group; the edit history is full of blues and reds cyclicly rejecting each others' changes, and collective progress is made fitfully, if at all. This step is best executed by a small core team for each point. Passive participants at this stage are selectively expressing their approval of statements, and this information is available in aggregate to the editors.
2) narrative construction and storytelling. Editors use the output of the thought circles to build an editorial. They are not allowed to change the words, although some expression is possible here by means of juxtaposition and choices about sequencing, inclusion, and omission. Again, a small number of skilled editors can fulfill this role. Passive participants at this stage express selective approval of the assembled articles. Based on data from the group formation stage, automated link structure analysis enables participants to easily find editorials they are likely to approve (or disapprove) strongly.
3) mind share visualization. Readers explore the space of competing editorials, annotating the constituent statements with approval or disapproval. As the reader provides this input, she navigates the space of editorials; with each new rating, an alternative editorial is displayed based on an automated search for the closest match from among available editorials.