Wouldn't our evaluation depend a lot on your model of how invites would spread across the social graphs of users? I'd think that the user base would expand unti

Wouldn't our evaluation depend a lot on your model of how invites would spread across the social graphs of users? I'd think that the user base would expand until the sub-marginal user had been invited. Given that the value of the network is largely determined by the user base, that user might be found fairly early in the invite process. Given that G+ probably had a heavily concentrated set of initial invites, I think it's a safe bet that it's largely done growing for now. That's probably good for Google, though; I imagine it makes the feedback better for this stage of development while being fairly easy to handle.