I did say in the post that:
The actual end result would be somewhere in between the tables below and the current witness ranking tables visible on steemit.com and elsewhere.
I could produce a list that is somewhere in between these lists and the current live data list - but for me the lists on this page are still valid as they show who are the 'movers' in the list, to an extent. I fully agree that the end result would be hard to predict if pushed to the live network.