Also listen on throttled writes/reads and increase to match

[felixsand]

We use dynamic-dynamodb to auto increase / decrease a DynamoDb table that handles sessions and sometimes (usually due to certain types of DoS-attacks) we receive sudden spikes that result in throttled writes. It would be wonderful if dynamic-dynamodb could listen on these CloudWatch metrics as well and increase the capacity until the throttling has subsided.

Maybe a new config value so that it's possible to scale up X% or X points if these metrics are triggered?