But, worded that way, your point is going to get you into difficulty if you try making that argument with a liberal who fancies themself trained to handle data, and I'd prefer it if you weren't caught off guard like that. You might still waste time clarifying with me, but at least I'm on your side.

Yes, comparing good thing of group 1 to bad thing of group 2 is as disingenuous as comparing bad thing of group 1 to good thing of group 2.

But comparing atypical good thing of group 1 to atypical good thing of group 2 is also fraught with problems, even though—and especially because—it looks like it's legitimate, since it can be described as a like-for-like comparison. This is, for example, how Dubai tries to gloss over its problems: by playing up the fact that it's a modern city just like any other.

The flaw of Art Candee's comparison isn't that it's disingenuous, even though the point (jingoistic supremacy) being made with the flawed comparison certainly is. As far as the comparison goes, the real problem is only that the samples aren't representative.

And comparing modern things enjoyed by a minority in two different places is also unrepresentative. It's very much a false equivalence, every bit as much as comparing terrible things suffered by the worst off in two different places.

To stick with the specific example of Vietnam and the US, these two places really do have genuine differences, as comparisons of representative samples will demonstrate. To compare them on a like-for-like basis, you have to pick and choose your samples and, in doing so, you introduce selection bias, undermining the validity of the comparison. It suggests a sameness that obscures the broader reality.

It's not that you can't, or shouldn't, point out the successes of various groups in their defence. I agree that doing so is important to counter the opposing narratives. But when you do, you shouldn't call them "completely appropriate" "like for like" comparisons because, statistically speaking, they aren't appropriate at all.

TL; DR

When out in the wild, avoid defending your counter-comparisons as "like for like". Instead describe them as highlighting the selection bias of the original claim. You'll keep a couple of gotcha libs at bay that way.