My point is orthodontia has a large market because, outside of actual pain and health symptoms associated with malocclusion, we have a predisposition towards well-aligned jaws and teeth.
I'm not making any assessment of the effects of orthodontia on long-term selective biases in human populations.
... but also, four or five generations is not insignificant. Evolution is just a change in allele frequency in a population over time, and five generations is more than enough time to objectively measure a change in frequency with such a large population sample. (I think the better argument would be that that the number of people that have had access to medical tooth alignment is a statistically insignificant fraction of the global population.)