You know, it may be that you and I are just too uninformed to make a solid case, either way.
For instance, a while back I read one of DHT's posts. She was talking about neurotransmitter research as related to the psychological types. I don't remember the details, but if I grab the idea and run with it, I see definite possibilities. For instance, one way to identify an introvert could be to measure their serotonin levels in varying situations: around groups of people; with a close friend; alone in their room. That's at least a start.
Socionics is also significantly more scientific than MB, I hear. You can probably ask someone like bionic about that, but there seems to be research involving physical appearance. I know from your prior posts you don't believe in it, but maybe it's worth looking into, or will be.
If you reject MB on the basis of it being "subjective," you're also rejecting much of what, currently, is psychology. A lot of the psychological concepts I've learned about from books, etc., couldn't be at all be quantified, per se, or verified by chemical tests or anything like that-- and, yet, they can be very useful in predicting outcomes in the physical world. If a scientific model is no more valid than its usefulness in accurately predicting outcomes, then even something like MB can be said to be a valid theory. It can and does with fair accuracy predict outcomes in the real world. It's not infallible, but then, nothing is. That's why even in the physical sciences we have to develop different models for different situations, because not a single one will cover every possible situation. The best that can be done is to find some conglomerate of all of them, find a way to reconcile them, because they so seem to contradict each other.
So, I wouldn't rule it out completely. It can be very useful and even accurate, whether or not there is a blood test for INTJ.