They did the math actually on that one - comparing smoking rate and obesity counter-correlations and the rise and fall across various countries. The smoking drop was modest, less than 10% which makes it a fairly clear net harm unless one defines net harm ghoulishly such that dying soon after retirement is a good thing due to reduced healthcare costs.

What an insane point of view. Both are terrible things that we should find solutions to.

"Both are bad" isn't an interesting or useful observation when there's a tradeoff between the two.