When the number of people is small enough, it's not too hard to figure out the identity of supposedly anonymous people.

I've done that once in an anonymous chat group with about 35 people in it.

Hard to say but my impression is that most academics are honest and would try not to do this, but also there are rivalries between labs and that tends to encourage "everything they do is bad and we're great" mentalities so it's definitely not surprising.