How do you know that the <title> tag contains the "original" headline? It seems to be more tailored for SEO, i.e. to attract the people who wonder (and then Google), "Who charges the electric scooters?"
For a popular article that will earn a lot of links its also a good strategy to change the title multiple times during its viral phase.
Since many people/cm systems link with the title tag it means you get a wider mix of anchor text which can help the article to rank for a more diverse set of key phrases
Not sure what content-management system The Atlantic uses, but WordPress (or at least its popular SEO plugins) allow you to specify different values for <title> and article <h1> headlines. A lot of news orgs are now savvy enough to use varied titles for different formats. For example, this week's New Yorker print issue has a story titled, "Behind the Wall", which works in print because your eyeball can scan the article's entire presentation and see the headline in context with a photo and subhed "As the U.S. abandons diplomacy, an Ambassador resigns in protests".
In the online version, the story's title is "The Diplomat Who Quit the Trump Administration", and the subhed (which is also the <meta> description) contains proper nouns: "For John Feeley, the Ambassador to Panama, moral failings at home seemed to compound tactical failings abroad"
The webarchive supports that this has always been the title.
https://web.archive.org/web/*/https://www.theatlantic.com/te...