It's just an email address. It could be Scott or it could be someone else. Yes, common sense would say it's Scott, but the reporter would have to still prove it's him. If you claim to be Scott, too, that will also need to be checked out.
Many people will take that information and run, but if you're writing for a national outlet, where accurate reporting is everything, your editor will say, "Yes, that might be Scott, but how do you know? What proof can you provide? If we get called out for a fact error, can you refute that claim?
The Scott who posts at slatestarcodex.com is the Scott who is scott@slatestarcodex.com.
Therefore, the material Scott when attempting to pin down Scott in the context of slatestarcodex is scott@slatestarcodex.com.
Human X out in meat space could or could not be Scott, but that much is immaterial, as scott@slatestarcodex.com has been shown to be directly linked to Scott Alexander the blogger as a means of contacting him.
Thus I ask: what better proof could one have that scott@slatestarcodex.com is Scott Alexander, author of slatestarcodex?
From an editor's point of view, that's not enough, assuming the reporter has not done any form of reporting through interviews, public records and other methods.
The story is about the blog, yes, but a portion deals with the _person_ or _people_ behind it. And that's important.
I'm not sure why though, the NYT, would need to know the name that is purposely never used.
If you really need the name sooo bad, then just don't dox him and drop the article. That's perfectly fine.
As long as they don't dox him everyone is cool.
If they can't write the article without doxxing him then they should just drop the article.
Whatever they do they shouldn't dox him. And if they can't write the article without doing so, then they shouldn't write it.