The evidence for cardiovascular risk factors and depression is intriguing, but not substantial enough to make strong recommendations for high-dose supplementation. Adverse effects are few and far between.

The ratio has an okay amount of evidence, mostly for short and intermediate term biomarkers. The best papers are here:

http://content.karger.com/ProdukteDB/produkte.asp?Aktion=showproducts&ProduktNr=229515&searchWhat=books&searchParm=toc

The health effects of omega-3 are less conclusive than most assume. My workplace was assigned by the federal government to do the omega-3 systematic review a few years ago. The results are summarized here:

http://ods.od.nih.gov/factsheets/omega3fattyacidsandhealth.asp

With regards to omega 6, there is much less research than there should be, because YOU CAN'T SELL A PRODUCT REMOVING OMEGA 6 FROM YOUR DIET!!!

In my opinion, this is a perfect area to apply evolutionary reasoning. We're never going to get high-quality randomized, double-blinded controlled trials comparing effects of different ratios. Since EFAs have more biological impacts than most any other single nutrient, it would make sense to do what has worked for 200,000 years.

Note that, while you can change serum markers of omega 6/3 ratio quite quickly (within 30 days there can be a dramatic shift), it takes much longer to squeeze out all those omega 6s you've been building up through the years that's stored in you adipose cells, cell membranes, etc. The reason this is important is that trials would have to follow you for a long time to fully ascertain the health impact of different ratios.

I would be very interested to see studies controlling for gluten/wheat intake though, that is an excellent point that Stephen Guyenet similarly made with regards to the China Study.

The amounts of omega 6 and omega 3 fats consumed do appear to have an effect on health and they do have important interactions with each other.

However whether the ratio or to the abosolute amounts of each is more important is I think less clear from the evidence. This difference is a little hard to explain.

If someone eats 0.5 grams of omega 3 and 20 grams of omega 6 this is probably unhealthy. But is it unhealthy because of the specific ratio is imbalanced or simply due to a deficiency of 3 and excess of 6.

If it is only the ratio that is important then eating 20 grams of omega 3 and 20 grams of omega 6 would be fine. Also only eating 0.25 grams of each would also be fine. I do not think there is evidence for this.

If the absolute amounts are more important it could hypothetically mean the following: That good health requires an intake of 2 grams of omega 3 fat and 10 grams or less of omega 6 fat per day. In this case the ratio of 3:6 will always be between 1:1 and 1:5 but the ratio is only reflecting the correct intake of each.

It is also still possible that none of it matters as long as you get adaquate omega 3 fats. The poor ratio in most western diets could simply refect omega 3 deficiency.

How these ratios also apply to different classes of fats from each group is also unknown. The omega 3 fat ALA (alpha linolenic acid) competes with the omega 6 LA (linoleic acid) for elongation enzymes reducing the amoujnt of LA converted into arachidonic acid while EPA and DHA do not. While EPA and DHA compete with arachidonic acid for conversion into eicosanoids.

It's all very complex and not as worked out and understood as many people like to think.

It might even be that the omega-6 that most American's get is all damaged O-6 in the form of processed seed oil.

I agree with the general tone of your question; in particular, it's far from clear whether or not Omega 3 supplementation is a good idea. I think a lot of the evidence is overstated. Getting some naturally from fish makes more sense.

What seems much more clearly beneficial to me is lowering Omega 6 intake. Stephan shows here that 4% of calories might be a good cutoff point of sorts, and there's evidence from the Lyon Study showing just that. Once you get your Omega 6 intake low enough, perhaps the 3/6 ratio becomes largely irrelevant.

I'd say that the issue is still far from resolved. In fact, Walter Willett, one of the world's leading epidemiologists, actually thinks that omega-6 fats are good for you:

http://www.pbs.org/wgbh/pages/frontline/shows/diet/interviews/willett.html

"And it was found that saturated fats increased our blood cholesterol, and polyunsaturated fats from liquid vegetable oils reduce our serum cholesterol. And so in the mid-1970s, the predominant dietary advice was to replace saturated fat with polyunsaturated fat. As it turns out, that advice was probably very good and had additional benefits beyond just those that influence serum cholesterol levels, in that [the] increase in polyunsaturated fat was probably largely responsible for the major reduction in heart disease rates we had during the '70s and early '80s in the United States. In fact, the rates of heart disease death went down by about 50 percent during that time."