Calling a spade a spade
How about:
- 1 – Sharing a story [of open care]
- 2 + 3 – Online conversation aimed at processing information
- 4+5 – Online conversation aimed at common action
Theoretical background again: Edgeryders has some stylized facts about hacker culture. But, again, I cannot provide any solid analytical method. I think you could apply methodologies from computational biology. They would work with agent-based models, where cooperative behaviour evolves from the dynamics of interaction across agents. But, as all models of this kind, you would need to take care not to predetermine the results by overengineering the assumptions. I will attempt some work on this sooner or later (with @mstn , hopefully) but not today. 
The abstract seems fine. Two possible improvements:
- Rewrite to smoothen up the English and approximate the style of writing typical of science, which is concise and snappy. For example, you list research hypotheses as pertaining to the paper, rather than to opencare. The referees do not care about the context anyway. If you do that, it becomes redundant to say that the goal of the paper is to disprove the hypotheses.
- Refer to data. As I noted before, data are very unique to opencare, because you have 200 informants (recommended minimum number of informants in ethnographic studies: 12), and because the primary data (posts and comments) and the secondary data (annotations and codes) are connected in the form of a network. This is a major methodological departure from run-of-the-mill studies, and the referees are likely to take notice. In practice, you could have a paragraph that says: we have this many stories, that many comments, authored by this many informants. We made X annotations using Y codes. We then proceeded to exploring this mass of data using [some method, for example I like to use the co-occurrences graph] as well as more traditional techniques of qualitative data analysis.