Thank you for the suggestion. It would be interesting to compute the Gini Coefficient. There is one concern though: in the case of wealth distribution of populations the people with very little to no money are the poor and they are just as real (certain) as the rich . However, in most token distributions there are many wallets with near zero balance that are most likely transaction leftovers.
For example, for EOS there are many many accounts with near zero balances (see on the linked figure). This large number is probably because of the many transactions causing rounding leftovers.
https://postimg.org/image/nga4g4yf9/
These would bias the Gini calculation significantly, thus we would probably need to introduce some kind of cut-off. But that would be tricky because the results may change depending on what cut-off value we choose. Let us know if you have ideas to tackle this.