I agree. For anything sophisticated you have to use the Kubernetes Python client combined with the KFP DSL. The compiled Argo Workflow spec can become such a me

I agree. For anything sophisticated you have to use the Kubernetes Python client combined with the KFP DSL. The compiled Argo Workflow spec can become such a mess that you quickly run into a size limitations -- especially if you choose to deploy a Python function directly (which then uses Kaniko to build the containers in cluster and run them via Argo Workflows). At that point I would rather use Argo directly. And KFP is generally quite behind in Argo Workflow versions. And it's not like KFP makes CI/CD (MLOps) any easier than Argo Workflows itself.