At some point we need to sit down and map, exactly what kind of information can be extracted from which meta-data sets. For instance, screen time is also sleep

At some point we need to sit down and map, exactly what kind of information can be extracted from which meta-data sets. For instance, screen time is also sleeping habits, for some its toiletbreaks. If its toiletbreaks, its stomach/health related. Couple that with location, its eating habits and risk factors. Couple that with phonelogs its much much more. Once we have that mapped out, we can do an honest evaluation of whats being tracked, because every extrapolation possible will at some point be turned into usable data (my theory).