That's not what the paper primarily shows. The gists are sentences in italic (in the section 3):
> Is HTTPS usage still growing? (Yes.)
> Does HTTPS usage differ by operating system? (Yes.)
> Is HTTPS usage equal across the world? (No.)
These sentences combined illustrate the difficulty of calculating a single accurate and representative measurement. Probably the following summary (mine) is better than nothing:
> HTTPS usage ranges from 40% to 70% when one looks at the ratio of page loads. The wide range is due to the regional, OS variations and different methods to determine the ratio. It is noted that Japan and South Korea had slightly lower usage rates, no reason is given though.
Not only publications like this, but all the contest on web. Just today I was reading Bitcoin related article and only realized how old it is when reaching the part where it promotes Mt. Gox as a trusty Bitcoin exchange service. It really annoys me and I usually skip articles/posts that are not dated.
