Glad it's doing better!
To be fair, it looks like that slowdown was due to an issue with Amazon's image-hosting servers, not our equipment. So can't take credit for fixing it - but technically wasn't our problem to begin with, either. At least, according to that traceroute.
(You can tell by looking at hop 6: that server's showing 22 ms responses, which means you can basically disregard any previous hop reporting more than 22 ms response time)
If the problem recurs, however, definitely let us know.