Comment by colechristensen

You are missing the point.

If you can predict someone's coffee intake based on testing of their microbiome then you've proven that coffee intake has predictable effects on the microbiome.

The important part isn't predicting coffee use, it's just the proof that there's you can predict and perhaps control in the opposite direction leading to more research.