I really want to see stuff like this tested in massive cohorts (ie. 1 billion+ participants) via variable dosing.

Here's how it would work:

* Companies that make products would randomly increase by 5% or decrease by 5% the dosage of each ingredient in a product headed to a specific town.

* The government would publish A/B groupings for every town in the US and every chemical they track.

* Average health data would be collected, and whenever a specific chemicals A/B group difference exceeds some threshold, investigation is done.

The main downside is production costs of products goes up, since every product will now have a final step of adding all the A/B adjustment doses to the bottle before sealing. But I believe this is worth it for rooting out chemicals that have low level yet very widespread negative effects.