I think this article will help to answer your question, and for bonus points, there is also a youtube video :)
That just left me with more questions! So apparently the nucleation sites and the surface area at the top (where gas can leave directly) are what drives the loss of CO2. Buuuuuut where's the balance? As in, if you've got a weird glass that's really wide at the bottom but small at the top (small surface area) will the nucleation sites at the sides of the glass be more important? Or a martini glass, that's hugely surface area-y at the top, with relatively little glass? For that matter, if you've got martini glasses (cone shapes) that are steeper or 'flatter', how does it influence it? When does it go one way or the other?