I think the idea would be to take the top X earners a month prior to the fork, and do the same the month following. Look at the min/max/average.
The longtail, seen in the Steem bluepaper, is unlikely to become anything else, but if the EIP has softened the curve even a touch, then that would be seen as an improvement i think.
Okay @abh12345, I can see that telling part of the story. You know all about (I don't) what is available, in the way of global data, so it may well be my thoughts on it are impractical ...