The smoking ban began mid 2007. So the only fair start date for comparison is 2006 as the last full year without the ban.
I'm with you so far.
I actually made an error because I quoted the figure for men. The actual figure for 2006 was 22%.
OK.
And the 2016 figure was 16.1%. Which is an even bigger drop of 5.9%.
So a drop of 0.59% / year over the 10 years (compared with 1.28% / year in the four years immediately preceding the smoking ban).
Although there were rises within that period (probably because of both a 'bounce back' from the initial drop at the bans introduction and the end of the recession) there has been a sustained drop amongst the most hardcore smokers. Smoking hasn't returned to 2006 levels.
You are ignoring the big change that happened in 2012-13 when vaping went mainstream. This has affected smoking prevalence far more than the smoking ban (unless you wanted to make a case for the smoking ban actually slowing down the reduction in prevalence, for which there is more evidence than for your position).
You're also ignoring the fact that as smokers become a smaller and smaller group significant drops in the number look less significant when viewed at population level.
No this was you when you posted CRUK's figures about the 20% reduction in the number of smokers and gave the impression this contradicted the prevalence figures I gave earlier (If there were only 10 smokers in the country ...) I'm pleased you've understood it now 
In other words, 5.9% of the entire population gave up after the ban. But 26.82% of smokers in total gave up after the ban. This compares to 21.4% 1996-2006, 14.3% 1986-1996 and 19% 1976-1986.
So your graph might make it look like nothing really happened post ban, but actually it encouraged the greatest ever proportion of smokers to give up.
These are really not the most useful figures though, as you explain above (as smokers become a smaller and smaller group ...) and as I explained in a previous post (If there were only 10 smokers in the country ...) and they are not the figures public health pay attention to. Public health focus on prevalence because they care about what proportion of the population are likely to die young or have poor health because of smoking. Your figures may be useful to prove a point to people who are not familar with this field but they're not useful for much else.
Can I ask why you're so invested in proving recent recent falls in smoking prevalence are due to the smoking ban in 2007, when all the evidence points to a post-2012 fall which is most likely due to vaping?