However, I'm trying to make this "whole is greater than the sum of its parts" argument. But, with Berkshire, almost all of its investments thrive well on their own. Buffett invests in each and every one because of fundamental reasons within each individual company.

My argument is that some companies are making a lot more money off of each traffic batch than the traffic batch is actually worth. By figuring out such companies, the value of the traffic to facebook would then be higher. Would this work with every company that advertises on FB? No. Could it might possibly maybe work with at least one or two companies? Yes.

Saying any company with a non-functioning business model should just do something else is really stretching it. I believe FB does have a functioning business model that could use a great deal of improvement. This, to me, could be such an improvement. What other alternatives are there? You have already said yourself that things like "social graphs" are just BS meant to inflate valuations to $15 billion.