^ This exactly. They may be accurately reporting the actual number of diagnosed cases (which I'm on the fence about), but they are certainly not reporting "suspected" cases. I understand that's hard to discern, but they should have a decent idea about the average mortality rate in the area, and be able to see an uptick that will give them a rough estimate.
Honestly I think it is hard job to get an accurate reporting giving the scale of the problem. We are talking about a single digit accuracy at the scale of 10k+. Even in H1N1 outbreaks, the medical visits and death tolls in the US was estimated with a wide band. That being said, the death of un-confirmed patients could be intentionally ignored to make the problem look not so big. But this is not a unique thing to authoritarian governments.
