1. I think a more universal content aggregation / discovery tool would be beneficial for many people. Google might be the analog that comes to mind. This makes me wonder how such an aggregator/search tool would operate in a decentralized way. Would each user/node be responsible for maintaining its own index? My question to you is how does this differ from current centralized aggregators and search tools?
2. Regarding distribution vs. publication: its an interesting point that you could separate the two, but I personally believe this benefit would be moot for most people since they would be unlikely to self-publish content. My view is that most individuals will continue to use centralized services, regardless of the underlying protocol, due to a combination of convenience, ease, and cost. Like you identified, unless there is some critical mass of people self-publing and self-distributing, I'm not sure I see a tangible benefit here.
Also, mea culpa, I sometimes forget that the author might read my response and perhaps my language in my initial post was too strong, I hope I did not offend.