fix: restore WHERE filter on pg_stat_statements alongside LIMIT 50

Both layers of cardinality control:
- CTE: ORDER BY total_exec_time DESC LIMIT 50 (hard cap)
- Outer: WHERE calls >= 5 OR exec_time_total >= 1000 (noise filter)

Root cause found: the deployed VM config was missing the WHERE filter
entirely, allowing all queryids through (~30K+ series).

https://claude.ai/code/session_01SzJxzZNQjDQphaHyaX3RU7

Author: claude

config/pgwatch-prometheus/metrics.yml

@@ -1037,6 +1037,7 @@ metrics:
           temp_bytes_read::int8 as temp_bytes_read,
           temp_bytes_written::int8 as temp_bytes_written
           from aggregated_statements
+          where calls >= 5 or exec_time_total >= 1000
       17: |
         with aggregated_statements as ( /* pgwatch_generated */
           select
@@ -1085,6 +1086,7 @@ metrics:
           temp_bytes_read::int8 as temp_bytes_read,
           temp_bytes_written::int8 as temp_bytes_written
           from aggregated_statements
+          where calls >= 5 or exec_time_total >= 1000
     gauges:
       - calls
       - plans_total