Answer
This points to the kubeconfig file for the Kubernetes cluster not having been correctly retrieved and placed on the Windows Management Server before running main.ps1 — the installer's PowerShell script and kubectl both need a valid kubeconfig pointing at the actual Kubernetes cluster on the Linux host to reach it, and without that step done correctly first, filling in either the blank or the Windows Server IP in the config just produces a different flavor of the same underlying connectivity failure. Revisit the server-preparation documentation step specifically covering how to fetch the kubeconfig file from the Kubernetes cluster on the middle-tier server and copy it to the Windows Management Server before re-running the installer.