Add filtering to avoid errors where the `namespace` the driver is being installed to matches one of the `ADDITIONAL_NAMESPACES` selected

[klueska]

We should add some filtering to avoid errors like this, where the namespace the driver is being installed to matches one of the ADDITIONAL_NAMESPACES selected:

 helm install nvidia-dra-driver-gpu oci://ghcr.io/nvidia/k8s-dra-driver-gpu \
       --namespace=gpu-operator \
       --version "25.8.0-dev-5483fa4f-chart" \
       --create-namespace \
       --set nvidiaDriverRoot=/run/nvidia/driver \
       --set gpuResourcesEnabledOverride=true \
       --set controller.containers.computeDomain.env[0].name=ADDITIONAL_NAMESPACES \
       --set controller.containers.computeDomain.env[0].value=gpu-operator
       
Pulled: ghcr.io/nvidia/k8s-dra-driver-gpu:25.8.0-dev-5483fa4f-chart
Digest: sha256:2ac2e0a4837064621b572453f7dcb999db043f6669a1a6623fd973668b42d707
Error: INSTALLATION FAILED: 6 errors occurred:
	* serviceaccounts "compute-domain-daemon-service-account" already exists
	* serviceaccounts "nvidia-dra-driver-gpu-service-account" already exists
	* clusterrolebindings.rbac.authorization.k8s.io "nvidia-dra-driver-gpu-clusterrole-binding-gpu-operator" already exists
	* clusterrolebindings.rbac.authorization.k8s.io "compute-domain-daemon-role-binding-gpu-operator" already exists
	* roles.rbac.authorization.k8s.io "nvidia-dra-driver-gpu-role" already exists
	* rolebindings.rbac.authorization.k8s.io "nvidia-dra-driver-gpu-role-binding" already exists

Originally posted by @klueska in #585 (comment)