I find that theory very problematic though, because A: it's not terribly necessary B: the government did not particularly jump on this as it would have if this was coordinated propaganda, and let's be honest C: if you want to make that case there's probably 25 equally large, true stories that could be published if the intelligence agencies were willing to burn a few sources or something as they seemed to in this story. It's not as if the espionage status-quo is "nobody's hacking anybody else, because everyone's afraid of the consequences if they do"; what public evidence we have suggests the status quo is more akin to a free-for-all.

Heck, just the HN replies to previous discussion on this story had some comments from people who had jobs for a while detecting espionage on their boards, which were created precisely because some was found prior to that. For instance: https://news.ycombinator.com/item?id=18138699 You could make more hay with that story than a faked-up one. There's no need to lie for propaganda's sake, because the truth would be more than adequate! (The propaganda element would be in the spin and claims about the scale of the problem, not the brute facts, which certainly exist.)

If this ever was government propaganda, they changed their mind before finishing the job and hung Bloomberg out to dry.

At the moment, the best theory that seems to fit all the facts, including the comments from people who were interviewed for the story, is a not-sufficiently-technical reporter very motivated to tell this story (perhaps because of the reward for stories that move the markets?) who compounded confusion upon confusion in a domain they were suffering from Dunning-Kruger until they were able to write that story. And presumably, similar problems in the fact-checking and editorial staff. I'm still not guaranteeing this, but I'm giving it at least 75% at this point.