What's getting you guys all confused is the tweet was originally posted while DST was in effect.
It no longer is. Time shown in archive now for EST is 18:15 [UTC -5:00].
This is why it's so critical to compare UTC timestamps ONLY when computing detlas (then convert back to a baseline EST) for making graphics.
The only reason we're running into this for this particular alleged marker is that it happened during DST. It isn't an issue for any others because they all happened AFTER reverting back to standard time.