@amoeba almost any form of closure of a migrated question will bounce it back as rejected (this is why it's imperative not to migrate anything that is likely to close). I think user603 has replied there though (which I guess you have now seen).

@Kodiologist It's a good and useful question, and on-topic at CV - and also at Math.SE. It's good to have around. But it should be closed as a duplicate. Unfortunately, none of the closing reasons apply to this situation.

(And I don't think we would get anywhere with a feature request for cross-site duplicate closing, this just doesn't happen often enough. If at all, we might be lucky with requesting an "other" reason.)

@StephanKolassa: stats.stackexchange.com/q/214770/17230 has some questions on specific metrics differentiating it from stats.stackexchange.com/q/214118/17230. (AICc & adjusted pseudo-$R^2$s being alternatives to cross-validation.)

Wrt stats.stackexchange.com/q/215369/17230, I'm not sure we have a policy on cross-site duplicates.

(I know we tell individuals not to cross-post, but that's a little different.)

I was going to say stats.stackexchange.com/q/31036/17230 was a duplicate, but the relevant parts are mainly in comments.

I've asked the OP to clarify what more they're expecting. On balance I'd lean toward leaving it open: it's well on-topic & a thorough explanation would be good for here.

@Scortchi: Can this stats.stackexchange.com/questions/120584 be a duplicate? I don't understand this topic well enough myself.

@amoeba: Not really, I think; it's more a sequel.

I would be chuffed if somebody thinks this answer is good enough to upvote, so I can mark this question as a duplicate (questions without an accepted or upvoted answer can't be dupe targets).

@Kodiologist: +1 and hey, I learnt the word "chuffed".

@amoeba Danke.