And Amazon, I guarantee you, doesn't have that much unused network capacity. That would be very, very expensive at their scale.
If it's expensive for Amazon to have multiple Tbps of unused capacity, imagine how expensive it is for any other provider. To match the absolute spare bandwidth of Amazon having only 1/10 extra network capacity, another cloud provider might need to have keep its network utilization at only 5%. Maintaining a network capable of serving over 20x your current utilization "just in case of DDoS" is bloody expensive.
[0] http://www.businessinsider.com/nobody-gets-fired-for-buying-...
http://googlecloudplatform.blogspot.com/2015/06/A-Look-Insid...
One major difference between Google and AWS is that Google will carry your packets between data centers on its backbone by default. Google will also carry packets as close to the customer as possible, whereas AWS will dump it off as quickly as possible.
So, even between Google Cloud and AWS it's not an apples to apples comparison.
Also, I'm not sure I buy the argument Amazon has small unused capacity in absolute value, relative maybe. They could also have 5% across 23 locations so if your application is distributed it can have even better resiliency.
I think you are seriously underestimating the size of Amazon's or Google's networks.