Halcyon, I honestly don't think any statistics existed anywhere a few years ago (when carcinogenesis would putatively been initiated) (although there likely will be soon) that gave such fine level of location accuracy that a person's residential location could be pinpointed so accurately, especially in the vertical direction where GPS systems, especially pre multi-constellation systems as were the norm until recently, have had accuracy issues. However, such precise data clearly has existed for a long time for the actual cells, due to the use of their GPSs as base stations to provide the AGPS ephemeris data. The combined effect of AGPS, raw-data capable (RTK-capable) handsets, fast update times (20 Hz) and so on however is now allowing the collection of extremely precise location data. (at least a few centimeters, probably even better, even when indoors, especially if you average it over time, and throw out the obvious cycle slips- (the outliers).)
So even though I am skeptical about the existence of such data now, in the future it definitely will be possible to retrospectively ascertain if a particular user has spent any appreciable time living right next to or in front of a cell site antenna (or any other health risk, also to determine their level of physical activity, strength and rate of their heart beats, etc.) and then price their health insurance, etc. accordingly.