Personally, I think that they'll probably see which of the third-party implementations they like and use it as inspiration (should the actually move forward with it). That said, their "radar" should pretty much just be "make the service reliable".
One could also make a compelling argument for not complicating a service that (aside from outages and other technical snafu's) has really done well *because it's so simple*. For that reason, I think adding groups would probably be a mistake.
Interesting to know that they're thinking about it, though.
