Even More Empirical Developmental Theory?

What if the interactions between the more fixed portions of the brain that correlate with things like emotion and general modules of the brain could be measured and compared between many individuals while they were taking exams meant to detect their developmental psychology stage (e.g Spiral Dynamics, Hanzi Freinacht, Integral, etc.)? A clustering algorithm might be able to group the data and see if there are brain connection clusters in the data that correlate individuals at different states, stages, codes, types, etc. in ways that are related to development.

How would we do it pragmatically - where we don’t assume we’ve already arrived at a perfect set of elements that map between the brain scans and the theory?