Sounds like something of a gamble.
There are a number of proteobacteria that provide benefits to your health. Most have heard of commensal strains of E. Coli (e.g.Mutaflor), which I mentioned has a role maintaining the lining.
Expansion of proteobacteria? I suspect this likely represents a favorable development, although as Sidereal mentioned, not necessarily. Obviously I would want to correlate this with symptoms...Has this guy experienced benefits as a result of his starch regimen? I think large doses of potato starch are probably not optimal, particularly for someone with known dysbiosis.
One of the key differences between commensal and pathogenic strains relates to the structure of the lipopolysaccharide membrane that surrounds these organisms, more precisely lipid A. This is VERY important. Lipid A is the really nasty part of the endotoxin. Its structure, in large part, determines the immunogenicity of the molecule.
Lipid A can be hexacylated, pentacylated, tri-acylated, tetra-acylated, etc. In this regard, the acylation state plays a huge part in determining why one organism is pathogenic and another is non-pathogenic and why some people demonstrate tolerance to the same organism when others experience debilitating symptoms and signs of an infection. Chemical modifications by both microbes and host, in response to many environmental factors, alter the molecular structure and in doing so alter the virulence of lipid A.
So, without getting too esoteric, those organisms that promote an anti-inflammatory response possess the chemical abilities to diminish the immungenicity of that molecule which stimulates pro-inflammatory response. These commensal organisms that we may need have the capacity to enzymatically deacetylate and therefore neutralize the molecules that cause so much harm by activating the host innate immune response. For example hexacylated Lipid A may invoke a host immune response that is 10,000 fold more robust than a triacylated molecule of Lipid A.
How do we get these organisms? These are the same organisms I have been speaking of: anaerobes, principally gram-postive, soil-inhabiting, t-reg modifying bacterial organisms, some commensal fungal organisms as well. Those with robust enzymatic de-acetylation of certain carbohydrate structures that comprise plants, trees, and algae do appear to be particularly important. I don't think this is surprising when you consider that these are the most abundant carbohydrates on earth.
The organisms that possess these special enzymatic capabilities are also commonly found in traditional starchy fermented foods. East asian fermented foods like natto and miso are good examples. These may represent adaptations to diets not as highly enriched with more substrate specificity than starch. So starch is good medium, it simply appears as if the state of the colonic intestinal microbiome in inflammatory disease (including ME/CFS) needs some other carbohydrates that are not so readily accessible.
I'm thinking a variety of very modest doses of highly-acetylated and sulfated structural polysaccharides, combined with a little resistant starch may be a prudent approach. I believe the variety of prebiotics appears to have the capacity to re-engage some of those anti-inflammatory microbes.