unless they have decided that they can't measure it over the pacific, so they'll just assume that it's down...
Saying something is down just because you can't monitor it is dishonest. I prefer to use accurate statistics. The number of confirmed faults was only 1 all year.
I can't help but think of my university IT department. It was well known that it took 15+ minutes to log on to university computers (even quad cores with 16GB RAM). At certain times of the week logging in could take over an hour. The problem had probably been reported individually by pretty much every lecturer at least once, as well as numerous students, with all tickets closed as "unable to reproduce".
The school of computing actually enacted an autonomy clause in the regulations so they could take over complete management of several computing labs not used by other courses, and install their own windows images that did not have the problem. It had apparently been that way for at least 2 years by the time I opened a support ticket myself "to add to the ticket count" and the ticket was again closed as unable to reproduce.
I re-opened the ticket with words to the effect of "Please walk to any lab, lecture theatre or classroom and try logging in yourself before closing the ticket", the response came the following week in the form of a university-wide email about a recently discovered issue with slow logins... I suspect their end of year report stated that they only had 1 confirmed performance-impacting issue.