There are a variety of strategies being pursued to make pharmaceutical products out of microbiome research. Broadly those are:
1: "bugs as drugs" - genetically engineered microbes that perform some function. The idea here is that having the microbe in situ performing some function will be a much better way to administer a particular compound (or set of compounds), or remove a compound, than a traditional pharmaceutical. Large pharma companies have studied phenylketonuria as a metabolic disease for which the bugs as drugs approach would be great. Find a microbe that consumes phenlyalanine at a high rate and administer it at high levels (or get it to stably engraft in the host) and you have a treatment that would be vastly better than current dietary regimes.
2: "community engineering" - this takes many forms from fecal transplant, to trying to engraft a certain small cocktail of strains, to altering what the in situ community is doing by feeding a probiotic. The idea here is that there are tens of thousands of metabolites that microbes are producing in the gut, and by balancing or tailoring the set of metabolites that are made, you can improve health. Fecal transplants have good data for clearing recurrent C. diff in phase III clinical trials - this is the best developed of the microbial therapeutic strategies currently. Everything else is phase I or before.
3: "microbial natural products" - this is the world I work in (shameless self-promotion - if you want to come work at a very early stage microbiome startup email me at will@interface.bio). The idea here is to find the particular chemicals/metabolites that microbes make that have positive influences on our physiology. Most research here is focused on immune conditions, metabolic syndrome/dietary stuff, though there is increasing interest in depression and other conditions.
At a broad scale, I would say it will be 1-2 years before fecal transplants receive approval as a therapy for recurrent C. diff infection, at least 5 years before a bug-as-drug will be available, and at least 7 years before a microbe-derived natural product is on the market as a pharmaceutical. It takes an incredible amount of work to get from these associational studies to a pharma-grade product.
In the interim, I think there will continue to be a bunch of diagnostics and probiotics companies that (IMO) are bordering on absolute nonsense. There is very little predictive value to the tests supplied by most of these companies, and the data on probiotic efficacy is bad in humans. There is good evidence of probiotic and prebiotic effectiveness in animal husbandry (e.g. fish and livestock) but the data just aren't there in humans.