I think of that component as a very good area for a startup, which is why I'm part of such a team. I don't think it's unrealistic for one person to use Twitter, Flickr and YouTube for their social networking, while another person uses Facebook and LinkedIn; both should eventually be able to share with each other through a service that specializes in bringing everyone together.
Does the user have to set that up? Or does it just happen automatically? That to me seems one of the main advantages of centralization of this sort of thing: users just automatically get publicly posted photos, posts, invitations, etc. from anybody they've friended and vice-versa.
With OAuth and login cookies, it's even easier now than it was when FriendFeed began, but still requires one click to add a service, and a clicking on "Allow" at the end site. Once that's done, a good product should be able to centralize the contacts and social data/activities.
So how do you handle the case of a non-techy user who logs in and wants to upload some pictures? They have to go to Flickr (or equivalent), create an ID there, come back, and link that new ID to the social network feed? Or do you end up providing that sort of service, too, in addition to the option of using an external service? If so, how is this better than a Facebook app that lets you feed your data from an external service into your feed?
At this point, the service would connect the photo sharing of different services to the user based on their choice (I use Flickr, so I share Flickr, or I use Facebook so I'll share that with everyone). There is more functionality than in a standard Facebook App, but Facebook apps (and open social apps, and iPhone apps, etc) would be part of the solution. The service is based on sharing from the services a user already uses, but there's nothing stopping the service from helping users discover new services that are perfect for what they want to do.
