As a complete amateur in these matters, I think of it as being surface area x excursion (i.e. the amount that the cone moves to & fro). If you think of the entire speaker cone moving forwards and backwards then the volume it occupies is a cylinder and that's something you can measure. While I typed that, Pete gave the same answer.
Thing is, very few manufacturers publish excursion figures, and you've got this thing called "efficiency" which incorporates the idea of not losing too much power to heat (through friction) so I never feel that I either can work out displacement properly or estimate potential volume either.
What I can do, though, is a quick mental calculation when comparing cabs. Since the only variable that I know for sure is the radius of the speakers, I can do a quick & dirty comparison of, say, a 1x15 -v- 2x10. Radius squared for a 15 is 56.25; radius squared for a 10 is 25, but there are two of them so that makes 50. So if someone tells me that a 1x15 is about as loud as a 2x10 that doesn't upset me.
If someone tells me that their 1x10 (that's a 25) is as loud as someone else'e 1x12 (that's a 36) then I expect them to explain how they've achieved that ... more efficient speakers, greater excursion than usual, cab filled with magic smoke, whatever.
And chees has nothing to do with it.