There is another specific reason why I make these comments here. When I see PWME posting sceptical comments about speculations about this sort of stuff on PR they tend to get shouted down (or told to push off) by those who want to speculate, sometimes abusively. It is easier for me to put sceptical comments because being shouted down does not bother me.
In the case of MTHFR there's a lot of solid research into the missense mutations which cause reduced enzyme activity. While many SNPs (especially non-coding ones) on many genes are prone to having vague, contradictory, and low-quality support for their impact, that is not always the case.
It's been found pretty consistently that MTHFR C677T (rs1801133), for example, results in 70% reduction in enzyme activity when homozygous, and about 35% reduction in enzyme activity when heterozygous. Numerous studies consistently associate this with a modest elevation in various health risks: elevated homocysteine, and various folate-related birth defects when the mother has a mutation.
Things can and do get a bit crazy at that point, and some people then engage in various histrionics about their "broken" genes and the need to spend a lot of money on low-dose Yasko supplements which ironically don't even take into account the reduced ability to convert inactive (and cheap) folic acid into an active (and expensive) form. But again we can go back to the research, and find a couple of papers which show that people with these mutations completely remove the elevated risk factor if they supplement with an active folate or simply eat a respectable amount of vegetables as part of their regular diet.
The other common source of passionate misunderstandings is that people assume that having one of these mutations is exactly equivalent to having a disease, and that they are "diagnosed with having MTHFR" or similar. But if we look at prevalence rates of these mutations, it's obvious that they're far more common than any disease (or all diseases), and people usually live with them with little or no problem. On average, people have a 30-40% reduction in MTHFR gene function compared to optimal function. Hence it's nothing to get excited about when someone joins the 10% of the population with MTHFR C677T +/+. Just take a supplement or eat some damned vegetables, especially during pregnancy. Problem solved.
Oh, and methylation mutations don't cause ME/SEID. People just need to get over that, and move on with life. The research thus far shows no association, and that research is of a higher standard than a clinician saying "I see a lot of these mutations in ME/SEID patients" precisely because 1) they are not actually keeping track, and 2) they aren't seeing how ridiculously common those same mutations are in the general public.
I think there is a lot of quackery involved in SNPs currently, mostly originating with Yasko and those who mindlessly parrot her, but also branching out now with other profit-driven efforts generally run by people with no clue of what they're doing, and liberal cut-and-pasting from everything on the internet except the actual relevant research papers. But this doesn't mean that the entire field is worthless. It just means that a lot of misinformation needs to be corrected, and hopefully replaced with something more substantial.
By all means, please fight the misinformation. But there are some valid and relevant claims regarding SNPs which can be helpful for people to know about.