Reading some of the comments, I am reminded of the Greek concept of hubris...

In particular, it is funny how programmers, who make their living by controlling complicated systems, jump to the conclusion that every complicated system is trivially subject to human control. I had a professor once put it this way: disease is just that -- a dis-ease.

The first thing you have to understand is that no organism lives in a biological vacuum. Every organisms interacts with other organisms. When first two organisms meet, the interaction is usually rather messy. One or the other or both die in great numbers. Gradually, they make adjustments and the deaths decrease. One becomes a disease or parasite of the other. Eventually, the two organisms will reach a détente and begin living as symbionts. Given enough time, the distinction between the two may even vanish (see: endosymbiont hypothesis).

The flu virus is, in particular, a rather striking study of this process. Consider the vast majority of fowl who carry the virus asymptomatically. For them, flu is not a disease, but rather a simple hitch-hiker. For humans, you can predict the severity of a flu outbreak based on how well adjusted the virus is to humans. The more avian characteristics it has, the worse the symptoms and mortality are likely to be. At the same time, an ill adjusted virus is much less likely to spread.

One other thing to consider is that the flu virus itself does not exist in isolation from itself. The Spanish flu of 1918 is an interesting case in this respect. It was, as I alluded to before, ill adjusted to human hosts. At the same time, it was particularly communicable; something of an oddity for a flu with so many avian characteristics. It did kill a large number of people, but within a year or two the major damage had been done and the pandemic was over.

Why? Well, certainly a large portion of the human population that had become exposed but had not died were now immune. More importantly, though, the strain mutated and became better adjusted to human hosts. This strain, H1N1, then became the predominant form of the "seasonal flu". It is likely that you have become infected by a descendent of this very strain, probably many times over.

I would not fear this virus for the same reason that I do not fear Ebola or Marburg virus. Viruses that kill quickly and efficiently do not spread as well as those that cause some disease but allow their host to continue functioning more or less normally (all the while exposing many more to the virus). Of course, this is little comfort to the dead, and there will always be those who die from any outbreak (just go look at the annual death toll from "seasonal flu"). It is far, far from a reason to start predicting the end of humanity.

One final note on the bioweapon/bioterror angle: if you follow the thinking of the people that actually contemplate the use of bioweapons, you'll find that bioweapons are only ever considered as denial-of-area or knock-down agents. That is, you can use a bioweapon to prevent your enemy from gaining or holding a strategic position. What you don't find is anyone proposing to use a bioweapon as a coup de grace. Those that study these things understand that any biological agent has an in-built time limit on its effectiveness.

Edit: I just realized I should mention that while I don't appreciate the sensationalism espoused by the scientific press, H5N1 is a very real, very troubling threat. It won't wipe out the human race, but it does have a good chance of setting humanity back a decade or two (not from disease, but primarily from the knock-on effects of people reacting to the disease, halting productive work and worldwide travel). In my mind, though, this is all the more reason to publish this research and, while we're at it, pour some more money into the worldwide flu monitoring network.