Please be incredibly careful before bulk deleting those queues.
From a statistical modeling perspective, if any of those dummy queues accidentally captured real interaction data during your network tests, deleting the queue will orphan the historical data. When I run my Erlang calculations for long-term WFM forecasting, orphaned interactions cause massive variance in our seasonal volume patterns.
Always deactivate the queue first, run a volume report to ensure zero interactions are tied to it, and then proceed with deletion.