Are you asking for genetic (allele) difference, or phenotype differences? I see no reason to care about the uniqueness in allele frequencies except for geneticists, scientists and doctors because they can link a genetic mutation/disorder or identify lineages. Other than that, there seems to be no realistic purpose to calculate off the complete human genome, and it's always changing.
If you know about the Hardy-weinberg equilibrium, it is used often along with mendelian genetics, punnet squares, for plants to determine the chances of a particular offspring. To calculate humans, we can't, because humans are never in equilibrium due to factors that constantly are changing (intermarriage, new bloodlines). Thus, we'd never know when and how to calculate allele frequencies.
Assume in 10 years, 10 people with blue eyes died and 5 people with green eyes were born, 20 with brown eyes were born. The shift in the frequency changes that number, making the ratio for potential new blue eyed individual slightly less. If they were all in equilibrium, like 10, 10, 10 for each, it's much easier to calculate.
Also, during meiosis, it's totally random due to independent assortment. For example, if there was 8 choices for one particular allele for the child, only 1 was chosen. The rest are trashed and not used.
Because physical phenotypes are what we see with our own eyes, for example two people with blue eyes won't aren't going to have the same DNA for the responsible genes. It's like two people took a photograph of the same thing, but in different locations. Maybe one camera was a centimeter away from the other one. If you opened up their jpg files in text to see the ascii code, it's completely different. No one, and there is no point to really caring for the text unless you're a scientist.