It would be interesting to see empirical data on William Edwards's hypotheses, e.g. "Questions are getting fewer and fewer quality answers. Moderators and 'close this' clickers are getting increasingly narrow." An analysis like this might be feasible with Stack Exchange's API: https://api.stackexchange.com/docs.