No one is denying the history of medicine (the good and the bad), it just isn't relevant here. Lying to the patient is still lying regardless of whether or not the patient knows you are lying to them. Don't get me wrong, smoking is a nasty habit with negative long term health consequences, but what is proposed in the original post (in decent detail) is unethical by standards utilized here in the US.
There are other things to consider. Once you convince yourself that there is an acceptable reason to lie to a patient, you start to walk down a slippery slope. It seems reasonable that in this example most (if not all) of us would agree that smoking is not good for your health. However, what happens if you come to another reason where you think lying is justifiable, but this reason is not shared by the rest of the medical community. This can be dangerous on hot topics like contraception, both legally and ethically (not to mention it is not fair to the patient). In addition, once this deceit is uncovered, there can be long lasting consequences of this dishonesty. For example (on a much larger scale of course), generations of Black American's do not trust the American healthcare system for what happened in the Tuskegee Syphilis Experiment. It has been nearly 40 years since this study ended and greater than 30 years since the Belmont Report was released, and African American health in the south is still feeling the effects.
For these plus other reasons, it is generally accepted that there is very rarely a valid reason to lie to a decisional patient about care options.
For those of you who still have to take medical ethics, most of it is just talking in circles for hours, but there are key principals that you should take away; one of which is not lying to patients.
