To paraphrase the study by Sonnenburg: It is theorized that a depletion of the microbial biodiversity caused by a low-fiber diet is responsible for many chronic diseases. Mice that were fed a low-fiber diet developed less microbial diversity, but if they were put back on a high fiber diet, their microbial community recovered quickly. However, and most revealingly, if they gave birth to a litter while on a low-fiber diet (and therefore with a weak microbial community) the second generation mice were unable to develop a strong microbial community regardless of diet.
The conclusion is that microbes already present in one's body can adapt according to diet. If a microbial species metabolizes fiber and the host does not consume fiber, the microbe will possibly enter a "dormant" (my invented term) state, but it can be recovered to full strength. Still, "dormant" microbes will not be prevalent in the birth canal to ensure transmission to the next generation, so that generation will not have that microbe species regardless of diet. Over generations, this could lead to extinction of entire microbial species.
THIS is the evolution that the article talks about, NOT genome evolution.
> In that time [decades] we'll start altering the general human genome, and likely remove countless inherited conditions out of it either directly or indirectly.
What's the rush? You're talking about altering the human genome, something COMPLETELY IRREVERSIBLE in a single generation. This article asserts that an enormous part of our bodies was barely beginning to be researched five years ago! The reason this hubris is so common is that a few parts of our bodies are relatively simple and easily understood: the circulatory, respiratory and other systems that exist as meat. On the other hand we barely understand the many systems that live in a complex state of delicate equilibrium: endocrine, immune, and microbial which seems to be apart of both of them.