One billion WebRTC devices cannot be wrong...or?

So Google estimates there could be  1.000.000 WebRTC devices by end of 2013. One could criticise this as overly optimistic, but the numbers could be big still.
But how many is enough? The answer could be seen in at least a couple of ways:
- A needs to contact a well-known B who may or may not have a WebRTC device. If not WebRTC, then via gateway (a lot more clumsy), sol this could be difficult
- A needs to contact some B, like in playing a game where one wants to play against someone. This is quite good scenario for take-up if one looks the Chrome and Firefox market shares
- the take-up will be determined by network effecs. Not Metcalfes Law, but Reed's Law. Again, one should think of the winning user scenario...