The Start php7.0-fpm service task sometimes fails with this message when you rerun the server.yml playbook after the playbook failed previously. If you haven’t already, you may want to update your Trellis to at least roots/trellis#682, which tries to prevent this. However, the output from your systemctl status php7.0-fpm.service doesn’t suggest it is the exact same problem.
If you can just start with a fresh EC2 instance, that would probably be the simplest option to try first. Otherwise, you could try the command below on your current instance, then rerun server.yml:
Using sudo, you may find some relevant debug info in the root user’s systemd journal:
sudo journalctl -xe