I just had an experience that illustrates the automated detection and appeals problem.
On a post by the America’s Cup on Facebook there was someone leaving links to a supposed livestream of the race on a dodgy site (despite the fact that it is completely freely available to watch on the official YouTube channel). These scams are fairly common with any livestream found on Facebook (I have even seen them on academic lecture series). I reported two links as spam: both by the same person with exactly the same content.
I just got the results back: one was removed because it was clearly spam. The other was not removed because “careful checking has shown that it does not violate our community standards”. Again, these are the exact same messages by the same user and clearly linking to a dodgy site with an .xyz domain.
I have appealed the second decision. I have no doubt the appeal will be denied.