You can track composition of scores from the pages of people you cite. For example, Adam Shapiro got 90% of his score from answers (above 2200). But not all of them matter, but only those which get up-votes. Probably down-votes also matter, but I do not know to what extent.
W. Willett has the highest impact point, but this brings him not the top score (above 56, but still well below the top achieved score of 351), because he does not participate in forum activity. This also confirms my hypothesis about logarithmic contribution of impact to score. Perhaps they did it for two reasons: a) not to give too high advantage to people with many publications comparing to beginners, b) to make it more honest - because some journals do not have impact points at all, but publication can be important.
You get some fraction from followers, but it is just fraction of a point, so it is not visible if you have score above 10 and about 100 followers. I also wrote in an answer to your similar question why your score from forum activity can drop across weeks. If one is passive and does not write, score can depreciate. It also might happen that you get down-vote (you have to check).
P.S. I know one more researcher with high score who is missing in your list by Mostafa: https://www.researchgate.net/profile/Kenneth_Towe (199)