But a advance lab to create some superbug in some super Crag Venter style is hardly credible. Using evolution to do the dirty work is even less so.
To give you some examples without just a list of reference to papers you probably can't get access too. Common yeast is not the same as lab yeast. Lab yeast does not clump like wild yeast. This is purely unintended selection over the generations with particular lab protocols. Cool eh. We can get some lab strains to replicate quite fast in some media at 37C. Only it turns out that a lot of the "optimal" temperature thing is based again on unintended selection on strains optimized for the most common default lab environments.
We are currently working on tamiflu resistance in evolutionary experiments. Yes we observe resistance in only a few dozen generations. But this already happened in the real world (yea tamiflu more or less never worked properly anyway). We also observe hitchhiking of many things that don't look too good outside the media used. In fact we have to control for the fact that things happen on this media. We let there be a few 100 generations on the media so we don't confuse what is just selection for that media.
And of course all that happened without the presence of a rapidly responding immune system. Now move it into a real host (different media) with a immune system, and you find these lab strains just die out. We see this with lots of different systems by the way.
Bioweapons are even worse than dirty bombs. The threat is not from a terror organization or a stupid scientist. But just old fashion evolution. These make poor bio weapons. But a new pandemic on the scale of Spanish flu would be expensive. However not as devastating as the movies make out. In fact many believe a disaster movie level pandemic is impossible due to the tradeoffs that must be made in the design of the bug/virus.
But don't get me started on these recent H1N1 or bird flu bullshit. I will rant all day.