Both Enge and I address both the definition of "active" and of "invisible" non-public activity.

1. We're limited to directly observable public actvity. Given the sitemaps providing a population of profiles to sample, that means looking at actual public posts. Anything else would be Making Shit Up, which I prefer not to do.

2. It's possible get some sense of non-public activity by looking at followers and views for profiles with and without public activity. Evidence is that profiles without public posts have about 4.3% of the activity of those with. That gives roughly 4% of the total G+ userbase, or about 95 million users.

Again: showing activity at any time. That about doubles the active users count in total. (112 million + 95 million). It doesn't speak to recent activity, though.

My initial G+ "public activity" analysis: https://ello.co/dredmorbius/post/nAya9WqdemIoVuVWVOYQUQ

Follow-up "active" public sharing accounts vs. "inactive" non-public-sharing accounts https://plus.google.com/u/0/104092656004159577193/posts/RhnK...