This is solved - you need to specify the proper endoint for the zone you’ve specified. In other words, if you specify zone: us-west-1c you need to specify ec2_url: https://ec2.us-west-1.amazonaws.com
Otherwise you’ll run into the detailed behavior.
A lot of folks have run foul of this lately. I’ll look to change the connection method this weekend, I think we should move away from the env variables and begin to explicitly declare region and zone. This also ties into some things we discussed about ensuring the aws-related modules are used for consistent deployment of the same infrastructure regardless of the region and zone.