Hallucinated citations are polluting the scientific literature.What can be done?

Look at the example in the article, which is a fairly typical citation: While you can replicate the title, how do you propose to retroactively publish a paper in a specific journal, in a specific volume, on a specific set of pages, potentially years in the past?

Moreover, citations are most commonly for other people's work. And since you would be more likely to catch fake citations for your own work, the proportion of those is probably greater for fake citations.

So the people who would have to accomplish this, would be an entirely different set of people than the authors who published the fake citation. These people may not even be working together regularly, but you would need to involve every named author, as journals do check this.

And what would their motivation be, to publish based on a title that is potentially nonsense? A single citation that may not even be picked up due to the inescapable differences between the fake and post-hoc real citation?

I can't imagine that anyone would find that worthwhile