I agree with you that Facebook usage patterns do not follow a power law, but he doesn't imply that all of Facebook's 500M users are directly connected to the su

I agree with you that Facebook usage patterns do not follow a power law, but he doesn't imply that all of Facebook's 500M users are directly connected to the superuser group. He doesn't rule out the possibility that many users won't derive value from the them. There could only be 100,000 superusers (although this strikes be as an obscenely low estimate too) but have only a subset of Facebook's engage with their content.