[Feature Request] Support configurable NQ for concurrent vector search

VectorDBBench sends one query vector per search request. Changing concurrency cannot
represent applications that send batched searches.

Add --nq (YAML: nq) to set the number of query vectors per concurrent request.
Default to 1 to preserve existing behavior. Start with Milvus and Zilliz Cloud
single-tenant vector performance cases, including filtered searches.

Expected behavior

• Keep the configured concurrency list and record NQ in results.
• Count successful query vectors per second; measure concurrent latency per batch request.
• Keep serial latency and recall at NQ=1, concurrency=1.
• Reject unsupported backends and workloads before data loading.

Example

Search an existing collection with two query vectors per request. The collection must
contain the selected dataset and a matching HNSW index.

vectordbbench milvushnsw \
  --uri http://localhost:19530 --collection-name VDBBench \
  --case-type Performance768D1M --m 16 --ef-construction 128 --ef-search 128 \
  --k 10 --nq 2 \
  --skip-drop-old --skip-load

YAML uses nq: 2. Omitting --num-concurrency uses the configured concurrency list.

No files or tests are named. Start by tracing the vectordbbench milvushnsw CLI and YAML configuration into the Milvus and Zilliz Cloud performance cases; verify the documented NQ, concurrency, latency, throughput, recall, result-recording, and early-rejection behavior for filtered and serial searches.