ACORN fixes filtered vector search, except where Qdrant's planner already does

Filtered Vector Search: What Acorn Fixes, and What Fixes Acorn

Filtered vector search breaks when metadata filters turn a healthy nearest-neighbor graph into scattered islands. Qdrant repairs that damage in two places: filterable HNSW adds extra edges at index time; ACORN steps through neighbors of neighbors at search time. On a one-million-point benchmark, ACORN pulls recall back at 2.1x to 2.9x the plain graph's latency, then stalls on the 1% filter. Qdrant's query planner, free to route each query, holds 99.9% to 100% recall on all filters.