In the end there are huge range of causes that all impact these statistics.
For example, the rise of video games which may be a stronger explanation. One way to test this is if their are dips in violence around the release of GTA or other ultra popular games. With the method of action simply being fewer human interactions.
That said, the indirect evidence is even low levels of environmental lead contamination are significant over a population. Which suggests other things may cause similar issues without people noticing. EX: mad as a hatter
The correlation matches the phasing out dates in various places. The dates lead were discontinued weren't the same everywhere. Edit: how would games explain the geographic distribution? Areas with heavy lead contamination seem to have more of the effects, although no mention is made of game console/computer prevalence in any of the lead articles.
ED: This is in terms of the New York City drop in violence around 1993.
The data is just not as clear as your suggesting. After controlling for various things there is a percentage of the variation that links back to lead levels, but that does not mean it's the only or even the strongest impact in the raw data. Further, lead usage is linked to other things and as such is not completely independent.
PS: Remember, people move so lifetime exposure get's really hard to track accurately without regular blood tests.
I'm struggling to find the other article that covers a broader geographic area, it was posted here last year. This one is a ok http://www.bbc.com/news/magazine-27067615
I think the lead exposure = more violence connection is pretty well established. Leaded gasoline, while being the most studied connections out there, isn't even the only study in this regard.
For instance, there's this rather extensive study that shows higher homicide rates (among other things) in cities that used lead pipes for their water systems in the early 20th century.