From what I see, Darwin stated the number in this post:
http://forums.anandtech.com/showpost.php?p=38347376&postcount=358
He didn't show the derivation of it, so I'm curious as to what buckshot's issue with this is.
It is fairly straightforward to calculate the volume needed. You'd first have to agree on what "flooding the earth" means. Does that mean 100% everything covered, even the top of Everest? If so then you'd get a HUGE number. If you agree with something like 95% of the land covered then you'd get something lower, obviously.
Once we agree to a "flood height" then we need to find the volume of a shell with an inner radius of the distance from the core of the earth to the sea surface and an outer radius of the distance between the core of the earth to the "flood height." Then subtract the volume of all land above sea level. All these can be estimated close enough to get rough approximations of the magnitude of the volume of water needed.
The volume of a shell is the absolute difference between the volume of the smaller sphere and the larger sphere.
I'm guessing buckshot's hang up is on what is considered "flood height" because the rest is straight math. I have not checked Darwin's math, but I may when I have more time.