Document the S3 assumptions in the README

[PaulDuvall]

People just running the Dromedary stack can use the publicly-available S3 bucket to run the stacks without additional setup (assuming we keep it up to date). But, if they want to fork or use the repo on their own, there are no instructions that explain the process they should use to do this. The bootstrap.sh might solve some of this, but we also need to point users to this.