It really depends on the distribution of people on this forum, the number of times they've posted since the reputation system was in-acted, and various "trades" of reputation points given.
As has been pointed out before, who has strong Fe is not necessarily easy to tell from the type. But I think you were referring to the "lasagna" model.
So you can divide the types into twp groups:
With Fe in top 4 from lasagna model: ESFJ, ENFJ, ISFJ, INFJ, ESTP, ENTP, ISTP, INTP
With Fe not in top 4 in lasagna model:ISFP, INFP, ESFP, ENFP, ISTJ, INTJ, ESTJ, ENTJ
IMO, the top group forms a bigger section of our forum than the bottom. We could count using self reported types, or by the poll, or the survey to get a semi-accurate number.
With above caveats (and many more I haven't mentioned), I will attempt once again to do math in public:
For now, lets us call the fraction of the group with Fe in the top-4 of the lasagna model, F.
Let us represent a person symbolically by "f" if they fall in the top group, "t" if they fall in the bottom.
Then if we choose 10 people, we will have something like:
fttftffft
The probability of finding that particular (order dependent) selection of people
is given by F^<number of f's in the pattern>*(1-F)^<number of t's in the pattern>
The probability that any 8 of a random sampling of 10 people on the forum is in the top group ("f") is given by:
<the number of different patterns with 8 f's and 2 t's>*F^8*(1-F)^2=
(10 choose 8)*F^8*(1-F)^2=45*F^8*(1-F)^2
If we say F=0.6. Then the probability is 45*(0.6)^8*(0.4)^2=12% roughly that we would have exactly 8 members.
But I suppose you wanted to know what the chance of finding 8 or more is.
You could add up the chances of finding 9 and 10 to the list.
So we would add: 45*(0.6)^8*(0.4)^2+10*(0.6)^9*(0.4)+(0.6)^10=12%+4%+0.6%=17% roughly
But of course if F=0.8
45*(0.8)^8*(0.2)^2+10*(0.8)^9*(0.2)+(0.8)^10=30%+27%+11%=68% roughly
The expected number of people in the top ten would be F*10.
Hope that makes sense. I'm too lazy to actually do the counting.