I think a better approach would be to make an experiment and set people at different times (say 3, 5, 10, 30 years) and see how they compare. Maybe a 3 year tri

I think a better approach would be to make an experiment and set people at different times (say 3, 5, 10, 30 years) and see how they compare. Maybe a 3 year trial is enough to draw certain types of conclusions but not others. In any case, maybe 3 years trial might be enough to justify in the future a longer and bigger experiment which might be the point of these experiments.