Wow, talking about missing the point. The idea is that the Myers-Briggs personality test is bunk because it has two options for each category, when people are actually along a spectrum for each category, and most people are nearer to the middle than the extremes. Right. That's actually why the Myers-Briggs scores themselves actually put people along a spectrum according to how strong each trait is, and if they're close to the middle they say to count yourself as X instead (at least the Keirsey one that anyone uses does this; I have no knowledge of the official test you have to pay for, but no one else does either unless they pay lots of money for it, and that's dumb given the free availability of the Keirsey one).
The MBTI type scale is a highly useful way of coming to understand significant aspects of other people's preferences and tendencies. The problems with it have to do with focusing on four different ways that people can really differ and not getting to some other ways people can differ, but that just means it's incomplete. The idea that it's bunk merely because the scales it introduces are a matter of degree reflects nothing but total ignorance about how the scales function (since they are indeed scales and not binary categories). It's sad that someone with the readership of Ann Althouse is helping to perpetuate such a ridiculous criticism.