We could try to invent more idiot proof protocols, but the more control (and centralization), the more it will be "a kind of Internet". Not sure the founding principles and factors that made the Internet successful would resist anymore.
Anyhow, you can think all you want about security, such as Apple with its concept of his locking iPhone, but you can’t anticipate the unexpected, like the ‘jailbreaker’ success ... and peoples acting on their routers with their feet and not with their brain and their hands.
It's the nature of Internet technology: something could always fail and the ability to prepare for the unexpected is one of the reason why it works and is so scalable. It's also why "best effort" / "real knowledge" / education are better approaches than searching for yet another killer secure protocol. But maybe I'm a dreamer :-))
Anyhow, I’m not saying that nothing can be done. I can see at least two possibilities:
1/ What measures can be taken to prevent such things from happening and great discussion about it on the list.
2/ How can we take a more proactive approach and be informed of such incidents as soon as they occur and not after the first customer complaint
On first issue,a lot to do ... If ‘best effort’ is something that always exist in the today business world, I think we’ll arrive at an equilibrium without waiting for geni.net (certainly good) answers.
On second issue, there are plenty of possibilities and it's not difficult now to be informed to the minute when "big destination / AS" seems to be in trouble.
FYI, just see:
a very interesting TCP traceroute yesterday, during the "mistake" on Youtube (seen from our system / AS15436 on US East coast, Europe (Paris and London)
: very interesting, there is an abnormally high response time, just before severe breakdown. Is Pakistan trying to announce more and more Youtube address and less and less Youtube servers available to answer ?