Thanks, this is more clear and straightforward about where they got that number, though the relevant citation ends up being the same as in the longer document.
The evidence cited in that summary is the study 'Estimating the Harms of Nicotine-Containing Products Using the MCDA Approach'[1]. If you read that paper, I guarantee you will find yourself much less confident in the 95% harm reduction claim. Briefly, here's the method they employed:
* Convene a panel of "international experts" in a two day workshop.
* Ask each participant to separately rate how harmful they believe each type of nicotine-containing product is, on a scale of 0 (no harm) - 100 (cigarettes).
* Weight/average the scores and publish the results
In case anyone is inclined to think that counts as "evidence", be aware that the authors of that study don't:
"A limitation of this study is the lack of hard evidence for the harms of most products on most of the criteria"
They fail to mention in particular the lack of long-term evidence for some of the products, which have not been available long enough to properly measure.
In any case, I am aware that vaping nicotine is much safer than combusting tobacco (though I think the latter is a lot yummier). However, I found this to be an interesting case study in how a report can pull out precise and scientific-looking numbers like 95% and trust that people for the most part will take that at face value without looking into the scientific validity of the claim.
[1] http://www.karger.com/article/FullText/360220