I would consider any question about the presentation of a scientific paper in the media to be on-topic. Examining if the conclusions of the paper are presented accurately in the media is perfectly fine.
When you are asking about a specific paper that has not been mentioned in the media at all, then we're certainly crossing into the research-level area and we should probably close the question.
I don't think we should shy away from critically examining scientific papers, even though it is problematic with our huge scope and the lack of experts in all fields. If we restrict it to papers mentioned in the media, the vast majority of what is left is going to be about common skeptics topics like evolution, alternative medicine and climate change. There are often skeptical bloggers and others writing about the details of those widely publicized papers, so there should often be enough material for us to work with, even if we don't have an expert in that field on the site. But we should be able to attract experts for the most popular field over time on the site, which should greatly improve our answers to such questions.
The CLOUD question is about a widely publicized paper and I'm seeing it more as the question if the media representation is accurate and also if there is any contradicting evidence in the peer-reviewed literature. I would regard this as on-topic.
The newborn mortality rate question is about a statistic that is very often used as evidence in comparisons of the US healthcare systems with other countries. I also think this on is on-topic, though I think it would be a better question if it asked whether this statistic means that the US healthcare system is worse than those of the other countries. I suspect that this claim is the motivation behind the question, but it is really a related, but different question.