Not some of the points, all of them still apply.
Ok, this is the same question, just with the terminology way abstracted, which still doesn't change the premise. The data was not meant to note the error of particulars. It was merely a fraction of the illustration of the fact that the theory is flawed. It doesn't matter how much you abstract or change the language of the question when it is the theory behind the language that is the problem.
First of all, “predominantly green-eyed,” what does that mean, that ¾ of the eye is green and ¼ is brown? Also, what are you going to define as “deviant/uncommon personality traits” and by which standard?
Secondly, you already disproved your own thesis. “Because the total percentage of people with deviant personality types is so low, it is highly unlikely that these would compose most (>50%) of ANY data sample.” But that's not true. For example, we could probably say that most artists are NF. Of course, artists can also be many other types but this kind of hypothesis could actually be legitimate. But we can't say, for example, that most black people are SP. That just doesn't make sense and, essentially, the theory behind such a statement is exactly the same as your hypothesis.
But I think what you might be wanting to know is whether the way that one is treated as a green-eyed person correlates to the development of INTX traits. This would lead us back to the age old nature vs. nurture question, to which I would reply nature.