Facebook's own automated moderation is all but certainly playing a role here. To think that this group requires
no moderation is all but certainly false.
There's also a distinct difference between total registered members and current active users. Especially on any group more than a few years old, there will be many registered-but attrited members who aren't regularly active. The active membership may be anywhere from half to 1% of the nominal registered count (and even smaller on very large groups). Unless you've got accurate insights into those patterns, I'd discount your experience strongly on these grounds as well.
Looking at Google+ groups in the months prior to that service being shut down, I noted a number of patterns. Peak engagements, indicated by visible post, comment, and +1 (equivalent to "like") activity seemed to peak on a per-user basis at around 10,000 users or so. Not only were there far fewer very large groups (1 million+ members) than would be expected, but those that did exist were far less active than their raw size would suggest. Post activity was a much stronger predictor than size for engagement. Spot checks of larger groups showed many were overrun by spam and low-effort posts. (This despite Google's own anti-spam and anti-abuse measures.)
If you're in but not moderating that large subreddit, I'd also tend to discount experience. Reddit also has its own automated anti-abuse systems, though AFAIU it relies far more on moderators, and the load and tools are a constant source of contention between the site and its (nominally) unpaid volunteer admins. Lack of trasnparency also leads to many suspicions and complaints (most unfounded, some not) against mods.