This is a question for the biologists. Now that we are learning more about the microbiome and the intimate relation between bacterial species and humans (and presumably other animals as well), I am wondering if the gap between the colonizer and the colonized is even smaller than imagined.

For example, can the host use its microbiome to outsource its own evolution? In other words, can I use my microbiome as a fast and safe sandbox for natural selection, and then import better genetic material once it has been tested? How would one model and test this hypothesis?