Earlier quoted context omitted.
If the business proposition was simply "make another Visa" then sure, the proposition would be uninteresting. On the other hand, 2-3% of that $4 trillion is being consumed by the payment processing chain. Only a portion of it goes to Visa; the rest presumably go to banks and other layers of the service. Facebook potentially could, by virtue of being a trusted holder of identity worldwide, subsume all layers of this c…
2-3% isn't a dead weight loss, it's paying for very valuable things, especially fraud protection. You can't just magic it away. Also, as I said, if facebook wants to step into the payments middle man role (which they could) they could maybe make as much as Visa does. If they completely reinvented themselves and turned into a global merchant bank AND managed to convince at least a tenth of all merchant account holders…
My point is that Facebook is in a position to try building a new system, and the payoff is potentially enormous. Having a definition of identity that extends beyond a name and a cc# goes a long ways towards limiting fraud.
However, I tend to agree with you as far as Facebook's technical competence. This doesn't mean they shouldn't try.
As far as the chicken-and-egg problem goes, FB might do well to bootstrap a payments system in lesser-developed nations where Visa & Mastercard don't have such a lock on the market. A truly global payments system stands a good chance of eventually winning.