I agree with the migration (and perhaps I was involved in it? I couldn't easily see tell the voters were on our end by looking at the question in its new destination). A few points:
1) Migrating a question from one SE site to another should probably not be viewed as equivalent to other types of closing. It is an attempt to give a question more activity, not less.
2) Similarly, I don't think migrating a question from site A to site B implies that the question is off-topic for site A so much as that it is much more on-topic for site B. We want questions to be recorded in places where they are most relevant, where they have the largest possible community to answer them, where they can be compared to other relevant questions, and so forth.
3) The question in question is one about statistical practice. By good fortune, we have an entire site for that. Let me remark that the Cross Validated site is not only or primarily for academic statisticians. It is a general site for questions and answers both about the academic field of statistics and its application in a variety of endeavors. (In other words, if I am not mistaken it is more like math.SE than mathoverflow.) It's better for questions like this to be asked on Cross Validated, in which the leading answerers are all statistical experts and in which the community as a whole is statistically savvy enough to up and downvote accordingly. In a similar way a question which required mathematical expertise -- rather than expertise with the mathematical community or profession -- to answer would be better asked on math.SE than here, even though there are mathematical questions which are of interest to academia as a whole.