Median, normal lead exposure for toddlers in the 1960s and 1970s in any urban or suburban area would be 99th percentile by today's standards due to leaded gasoline vapors (and lack of awareness about paint dust).

So the crime hypothesis is more about baseline level of criminality being higher throughout the entire leaded gasoline era and for a few decades thereafter. It's generally framed as social science based on aggregate trends rather than individual dose-dependent epidemiological hypothesis.