I think you are seeing a pattern where there isn't one.
In the recent question about babies and swimming, it wasn't closed for being unclear about definitions. As I just posted in a comment, before seeing this meta-question:
It was an unclear question because all of the references you gave denied that babies can really swim. None of them made the claim that you were skeptical about, so it wasn't clear where you were getting that idea.
I could have equally closed it as Off Topic, because it didn't contain a notable claim. That has now been remedied. The question is currently open.
There is a question in the comments about what it means to be a "natural" swimmer, and another one distinguishing between swimming and the diving reflex.
Why is this relevant? Well, I haven't answered the question, but I predict it is going to be a dull discussion of definitions - that babies cannot keep their heads above water, cannot move themselves through the water and cannot completely prevent themselves from drowning. They often make what appears to be paddling movements, they can hold their breathe/slow their heart which increases their chance of survival, and they probably have a high fat-muscle ratio which promotes floating. Whether that can be fairly called "swimming" isn't a question of empirical evidence. It is a question of what the various words mean. (To paraphrase Dijsktra's analogy: Deciding whether submarines can swim isn't terribly relevant.)
I base this almost entirely on the references provided in the question, which aren't very definitive, so I am keeping my mind open that someone will post an answer that doesn't match my expectations.
Meanwhile, the question about whether dogs are loyal was a train-wreck, and remains closed. It remains totally unclear what the question means. How would we empirically test for loyalty?
You did attempt to define loyal, but it didn't help the question. If you define a dog's loyalty by the fact it follows its owner, and you observe it follows its owner, then by your definition, the dog is loyal.
The link you eventually posted claiming it had the answer didn't seem to address the issue in the slightest.