Correct. There's no such thing as a service that never goes titsup.com.
The only important thing is to work out the likely frequency and duration of outages and based on this, whether it's worth the cost of having a fully redundant system using A N Other service.
Given the length and frequency of outages in this case, short of using it as a deadman switch on the nuclear arsenal I can't think of any service where this would be worthwhile rather than just accepting a manual workaround for brief and rare periods.
There's also another matter to consider. In my experience, redundant systems are prone to failure modes due to the automated redundancy paying silly Bs and thus going titsup.com even though the primary services are all running perfectly well. Adding unnecessary complexity in case of the sky falling in is almost always a bad idea.
Biting the hand that feeds IT
