How can inter/intra-individuality and meta-analysis be connected into a more realistic model for recommendations in sport science?

Recommendations are made for the "average" public, while at the same time there is inter and intra-indivuality.

How can this be overbridged?

In some views there is already a link between individuality and average recommendations, for instance in fatigue-recovery models.