Apparently this controller adds the pod readiness gates indifferent to wether or not a pod is actually targeted by a VPC Lattice TargetGroup. This is counter-intuitive and diverges from the way the AWS LoadBalancer Controller implements this and results in pods never being ready when they are not used in a service.

For the AWS LoadBalancer Controller to inject the gates a pod needs to fulfil the following requirements:

• There exists a service matching the pod labels in the same namespace
• There exists at least one target group binding that refers to the matching service
• The target type is IP

I would suggest to do something similar here.