Explanation (tl;dr):
In this study, author is showing what is the difference between just balance and importance based harvesting. Technically importance is calculated on the account activity base (so transactions with over 1000 XEM). The result from this particular study says that actual NEM algorithm (PoI) is not much different if there wouldn't be this "importance" part in it, and harvesting probability would be similar in both cases. So the most probability of harvesting a block is comming from a "balance" parameter, not exactly from "importance" parameter.
good compendium from this https://docs.google.com/document/d/13M3wbdjI0Nzx-OKAd0dRCfOYFA_oT9Ydg96DkiUBsDs/edit?usp=sharing
There is also referred to a sheet
https://docs.google.com/spreadsheets/d/1LMLgretbB0EsRc8rAnvZQm_PEhtTeMqcku-VlN6iBTM/edit?usp=sharing
where one can compute his/hers own statistics. Maybe someone can find more positive results...
Just looks like it rewards highly active wallets instead of wallets that make a few transactions a week.
I don't think you are supposed to gain massive importance if you just transfer some XEM to an exchange every now and then.
"looks like it rewards highly active wallets instead of wallets that make a few transactions"
Where does it show so?
The correlation between harvested blocks and transactions is only 0.13.
ofc the "active" may mean in PoI also the amount of sent XEMs, but that is not visible in that data.
BTW, where is used that "you" instead of passive expression. Sometimes "you" is quite distractive