A more rigorous statistical approach was taken in this paper.
If I’m reading that paper correctly, the authors developed a theoretical model of how smoking prevalence would be affected by a smoking ban, and tested that model using empirical evidence from the Scottish and English bans (which weakly supported the conclusions of the model). So they weren’t simply analysing whether smoking prevalence dropped as a result of the bans; they were modelling smoking behaviour to predict if it would drop in the event of a ban.
So it’s not so much a “more rigorous statistical approach” as a completely different sort of study.
The model did predict that people smoke less as a result of a ban, however.
However, there is evidence it would cause economic harm to an already embattled sector and it would impinge on individual’s freedoms.
Is there evidence of this? It must be difficult to disentangle the many other problems pubs and clubs are facing and identify how much economic impact they have today from the existing smoking ban (and what the economic effects of the outdoor ban would be – positive as well as negative, because a ban will encourage some new customers).
Impinging on individual freedom? We have many examples of individual freedoms that society has collectively agreed to limit. The freedom to smoke outside a pub is pretty low on the Braveheart scale.