Reason: Mongooses belong to the Herpestidae family, which is part of the suborder Feliformia (if wikipedia is to be believed)

Edit: And civets too

civet -> feline

And all Feliformia are implicated to feline

Source

I think that usually the umbrella tags are family or lower level. If we did implications at the suborder level we'd have to implicate bears, badgers, weasels, otters, seals, skunks, and a whole lot more to "canine", since those are all part of the Caniformia suborder

Or, to stick with your proposal for implicating all of Feliformia to feline: if you searched for feline and got this, would you be surprised?

Good point, mongooses would be pretty out of place in that search

And what about civets? Some of them look feline-like (post #153717, post #121389), but the majority resemble a rodent/mustelid/other (post #129752, post #28462)

Like you said, the safest thing to do would be leaving them as a separate subfamily/family

Tags are for what is seen, not for scientific data. A picture is tagged 'feline' if it looks feline. As an example, I believe there's at least one species of cat that is occasionally mistaken for otters, and it's possible that a picture of such a cat might end up tagged 'otter' someday. If it really looks like an otter, this isn't a mistag. It's a correct tag.

Tag what you see.