Some data instead of "I have a friend who..." anecdotal evidence would be nice. The author doesn't make a very strong claim as to why facebook ads don't work which is the whole premise of the article.
I've taken this approach with PayPal. I'd use them for person-to-person finances, but I've heard so many small-business horror stories that I'd invest the time to make a merchant account if I made a startup. It's not exactly scientific, but PayPal's not exactly flooding the market with data either
It should be immediately obvious that the specific example mentioned is going to have a very poor ROI. Who sells on Facebook?
The comments I saw about Facebook's big plan came in two flavors. Some said "Facebook is getting creepy. I want my privacy." The rest said "this will be awesome, you'll see."
This guy is saying, "yeah, but will they be able to execute?" You don't need data to ask that question, and we all need to know the answer.
And does this guy really think that Facebook is incapable or unwilling of running projections on a key revenue indicator like their ad buys? An intern with a spreadsheet would be able to look at their internal data and realize that all their ad buyers are falling off a ledge and never returning.