The original test started on 5/24 (https://reddit.com/r/changelog/comments/4kuk2j/reddit_change...) which is why the chart starts a few days before that.

It has been a month since then.

Yes, but that test only applies to official apps... Do we have usage data on those? Especially taking into account very good apps already existed and most users probably couldn't be arsed to switch to the official one.

Those official apps have lots of downloads, but don't be fooled: I bet most people downloaded them because of the free three months of reddit gold you got when you logged into them.

The 16 subreddits in the 5/24 test were for the Web UI (the primary Reddit platform), not the App UI, which was available to all subreddits at launch.