I work in banking. Everything we do, that's easier elsewhere, is harder in banking, because it's a bank (regulated, money, etc).

Everything Facebook builds has to scale to billions immediately. Could be a contributing factor to why it took them this long to do this.

That makes a lot of sense. Does raise the stakes quite a bit too, which makes the decision maybe a little bit difficult. So the engineers who make the thing happen are for sure smart. And the researchers who prove it should be profitable are also pretty smart. But the idea itself -- trying to take over some other market that they already have a foothold in -- is pretty obvious.

Why would it? They can do controlled rollouts if they want. Core design has to scale, but I seriously doubt they'd be planning for millions/QPS right out the gate.