Zog, I am a bit torn on your post. Thanks for the link to Ben Goldacre's site. It is definitely worth a visit. The points you allude concerning issues with a blanket announcement concerning processed meats are well taken, too, but you seem to have missed the point on smoking studies and have fallen for the tobacco company line concerning cause.
The reason it is so difficult to assign causality to smoking's relationship with cancer in humans is because it is typically unethical to set up blind studies concerning smoking using human subjects. So tobacco producers continued to push back every time a new study was done ("It's just a correlation. It was an animal study and doesn't properly relate to humans. and so on...) while simultaneously suppressing any evidence that may have impacted their bottom line - the opposite of ethical. I remember from stats class in college that the only way to prove cause and effect through correlational data is to also demonstrate that there was no other possible cause for the outcome and that this is almost impossible to do using observational data. Perhaps a better way of putting it is that smoking has not been proven to cause lung cancer in humans though there is very good reason to believe that it does rather than simply stating that it does not cause it. I have often pondered what the world would be like if tobacco companies had to prove that tobacco use does not cause cancer.
But you may puff away until you are ninety and die of something else, cancer free.
So essentially you are saying that just because something doesn't kill you outright and something else might get you first, there is no causality? I apologize if I am putting words in your mouth, but the analogy that comes to mind is that of playing slots in a casino: the only way to not lose everything is to cash out early or simply not play.
As an aside, I realized the eventual source of my demise would come in the form of smoked foods when I heard they were found to have a link to cancer.