The analysis I would like to see (and really, who better than you to do it?) is the average number of shares it take for a post to propagate through, say, 90% or 95% of the community. My guess is not that many. I think its something you could easily do with the analyses that you've already conjured up.
The problem is data scraping is I had an excel file of the graph at a given date I could do it.
Indeed. As I've mentioned before, scraping data from Hubski would be awesome! However it's quite tough to do right now since there's no API and lots of the actual userdata is retrieved and viewed dynamically via javascript (not served on static pages). I've been thinking about trying to get some kind of basic data gathering tools together for Hubski. I really need to set aside a weekend and make it happen :)
the followers of the top 30 most followed users is around 1550
so about 1/4 of the total accounts. If I had a excel file of the entire graph I could do a distance and centrality study. from messing with the thing I think it is quit shallow but deeping.
between 4 and 6 shares.