Is your feature request related to a problem? Please describe.
This feature request addresses a user experience issue. While I understand this relates to Kubernetes' native orchestration, the current behavior creates a confusing user experience.

The current implementation also leads to poor error handling: if a customer mislabels a node, the controller still creates a nodeclaim but eventually times out because no controller handles the nodeclaim. This timeout error is disconnected from the actual root cause (the labeling error), making debugging difficult.
Describe the solution you'd like

1. Have the workspace controller automatically label the node when a preferredNode is specified.
2. If option 1 is not feasible, at minimum return a "no node found" error immediately when a preferredNode is set but no matching labeled node exists.

Describe alternatives you've considered

Additional context