I'm going to agree with Gerontocrat and call the minimum at 2,424,731 square km. Which means the eight people who guessed 2.25 -> 2.5 and the one who guessed 2.375 -> 2.625 have all proved accurate. Better luck next time to everyone else, although those who went for 2.125 -> 2.375 shouldn't feel too bad since they were, of course, damn close.
As is often the way in this forum, the median and modal predictions fell a fair ways south of the actual figure, both being in the 2 -> 2.25 band. Still not wildly unrealistic, though, imho.