That was my understanding too, but I haven't come across such evidence except when accompanied by yet more compelling criticism of it. I would love to learn of compelling evidence.

So far I have only come across cases where someone made up a model, started using it, and people followed, same as the And model here.

Anyone can make up any labels, attach behavior to them, and lo and behold, people will fall into one or the other category. Add a few years for people to get used to them and people forget the labels were made up. Add a few more years for the labels to make it into standard language so people grow up learning them and they feel like parts of nature.

Penn and Teller could probably go to town on the concepts on their show.

I'd love to learn something new and be shown something I missed.

FWIW, I'm not a psychologist. I only studied it for a few years.

Personality theory is one of the weakest branches on the tree of psychology because of its historical roots and because of the difficulty of applying the full apparatus of science. There is no universal consensus for personality theories as there is about, for example, behavioural conditioning. Not even close. My personality textbook was very, very thick and nobody mentioned in it agreed with anybody else.

The strongest results in personality work so far come from aggregating work from different existing models; that's how the Five Factor Model (including the Intro/Extra axis) was developed.

The strength of the FFM is that it comes from multiple lines of evidence (lexical studies and factor analysis of existing personality tests regardless of underlying theoretical model). That is: it arose from what appears in the data, not from what appears to fit particular theories of personality. Indeed it finds support from work done on theories that flatly contradict it. Put another way, the FFM finds widespread support because it is merely descriptive.

Eventually, as I say, more work from the neuro direction will begin to tease out what, if any, correlations can be found with underlying neurophysiology. A quick squiz at Wikipedia shows some early fMRI studies with different brain activation patterns for folk identified as extra/introverted.