My own figures are as follows: assuming I typically eat 500g+ of lean, cheap ground beef per day, plus butter up to my calorie limit, that gives me a bit less than 5g omega-6 per day (3g from the butter and 0.7g-1.4g from the meat). Obviously if you're eating other meats then you might be consuming a lot more e.g 25-30g a day if you're eating pork/lard.

If I'm eating eggs (typically 12-18 a week/2.5 a day) that's another ~12g o-6 a week/~2g per day. So in total, assuming 5g per day, I'd need one large can of salmon per day to balance out my omega3:6 ratio. In terms of fresh farmed salmon, (2.5g o3: 1g o-6) I'd need around 2kg of salmon a week to balance out. Based on the numbers of the packet of my salmon from the supermarket, the ratio is 3:2 and 3g omega-3 per 100g and so again I'd need more than 2kg to achieve a balanced ratio. Of course some people seem to think a 2:1 ratio is acceptable, in which case you'd need half as much. The canned salmon is wild and thus virtually all omega 3 conversely, so the necessary can a day only works out at 1.5kg per week.

Since most of my omega-6 is coming from butter and eggs, it doesn't much matter whether the salmon (40g protein per can) displaces any meat from my diet or nut. However, if you were eating a fattier, omega-6ier meat like pork or chicken, then it would be very significant if oily fish were displacing your meat.

All this might well make you think that fish oil is the way to go. On paper, this is certainly the case. It's also cheaper for me to buy 2.5-3g omega 3 from fish oil (23p) compared to 3g from fresh salmon (and this comes with omega 6 as well) for 100p. Canned salmon is a lot better, but still 70p for the same amount of o-3.

In favour of the fish oil is its convenience and the fact that I find it a lot more palatable than canned salmon! Salmon or other oily fish of course comes with the advantage of added nutrients (although cod liver oil would have lots of advantages in this respect I suppose). Fish also has protein, which for lots of people might be an advantage, but for me is probably a disadvantage, I don't know where I'd fit it into my daily meals.

Of course the major thing to consider and the major problem with fish oil (and undoubtedly why I don't use it) is whether fish oil works. On paper it contains as much omega-3 as fish, but Matt Metzgar links to some pretty compelling evidence that it just doesn't produce the same effects. This isn't a finding that I like accepting, since I naturally much prefer to look at the totality of the nutrients and other compounds contained in foods and judge on that basis which are better, rather than hand-waving that 'real food' contains some unmeasurable and unidentifiable magic benefits. Nevertheless in this case, as in some others, there does seem good evidence that there is something missing in fish oil that one gets from fish. Of course, if the problem is absorption, it might be that this problem is remedied by taking the fish oil with saturated fat, or it might be that ensuring that adequate micronutrition is the problem, but we don't know. There is also a suggestion here though taken from Julianne citing Andrew Stoll, that omega 3 stops being effective as an anti-inflammatory if not consumed with sufficient antioxidants (it suggests vitamin E and C) to stop it oxidising. I don't know if there's anything in this (especially the vitamin C part) but it could be another explanation. If oxidation were a worry then that might seem to be a reason to prefer fish oil, since it's not cooked, like fish, but then we don't know if/what other constituent parts of fish keep the oil from oxidising (since otherwise fish would seem to be quite unhealthy, despite all the evidence to the contrary).