I think the relationship theory is indeed the one big strength over MBTI.
I also agree with the idea that type is about the structure of your mind rather than about your preferences, or whatever you happen to score on a personality test.

I have noticed that there are often quite striking similarities in both appearence, behaviour, expressions and gestures of people with the same type. Most people who know about socionics seem to dissmiss visual identification out of hand but in my opinion it is only natural that people who have a very similar personality could also have similar looks, because both the way we look and the way our minds are organized must be at least partly determined by our genes. It could well be that either the genes that determine our looks also influence our behaviour, for example by modifying the level and response to various hormones, and/or that the genes that build our faces and the ones that build our brains are usually inherited together because they are closely linked in our chromosomes. Of course this is just speculation but I believe that visual identification is one of the most fascinating and promising aspects of socionics. Any theory of personality types that aspires to be truly scientific must be firmly based on the biological foundations of human mind and body. I think in the future genome sequencing and/or brain scans will be the most reliable tests of personality type.