The Atlantic is running a major article questioning the received wisdom about flu vaccines and antivirals, for both seasonal flu and H1-N1."
That's what the article's point is! It's not saying "vaccines don't work" it's saying "they say vaccines reduce the death rate by 50% and the numbers don't bear that out. What's the real number?"
And that's a fair question. We know the virus isn't 100% effective, it damn near killed this girl: http://www.google.com/search?pg=q&fmt=.&q=dystonia+flu+vaccine
Neither though is anybody saying the vaccine is zero percent effective or universally toxic, what happened above is a rare edge case (but as an aside it would be nice to be able to predict when this was going to happen, this is a fairly *catastrophic* edge case).
But the examples brought up in the article do suggest there is substantive argument that the claimed 50% reduction in mortality rate is indeed in question, that's all.
Nobody's actually measuring people who have anti-bodies of a specific type, the data gathered is fairly meaningless by lumping a lot of things (rhinovirus, coronavirus etc) as "flu", also the cohort factor and related effects do have a demonstrable non-zero effect on the mortality rate.
So, it's not a question of is the vaccine useful or nor, more like a plea for more accurate analysis and gathering of the data in question.