Mod parent up. Absolutely.
Of course the big sites don't have redundancy for a big unfortunate event like this and I question the commercial viability of even adapting their achitecture to suit rare events. Is it really worth scaling like that for a rare event? Is it really worth re-addressing their systems architecture for events that rarely happen?
How much resource would that reorganising take? How much would it cost? We're in an economic down turn, do they have the money in the bank to address this when the money may be better spent elsewhere in their respective businesses rather than auto-scaling on big events?
What's the online ad revenue like at the time of a big event? How is that revenue impacted at the time the service experiences difficulties? Is it worth swallowing the cost of downtime?
I'm sure all of these questions may be going through the CTOs' minds at their respective channels/agencies right now. They should be.
Now I'm not disagreeing with Reuven Cohen's comments per say, I'm just pointing out that there's got to be a good business case for scaling to this level and that business case is going to vary from news channel to news channel.