There has been a response in another ticket that they're harmless markers that have been around for a while, but it's hard to tell if they actually went through the data to confirm it.
That’s not an official response and, honestly, I doubt they went through every single data point to check whether they existed before this week or not.
If I were an idiot planning to use a public product for classified opsec activities, the least I would do is try and make them blend in with existing data to reduce the chances of it being removed for just this purpose.
but either way, I do hope the traditional "Google support via HN frontpage" actually helps in this circumstance, too