The post is definitely correct, but I think the App.net scenario is a startup unicorn from which valuable conclusions are difficult to draw. There's very little that they did that others could repeat.