To begin with, the reproducibility of your sample is in question - https://mrheisenbug.wordpress.com/2014/04/24/dear-american-g...

The resolution from uBiome data is also 16s, and (likely) too low to yield good information.

There is no definition of a good microbiome. The research is still in its infancy. Attempting to 'improve' your microbiome would likely be a mistake. The upside is that you may have contributed by paying uBiome.