refactor: unify prometheus_scrape_interval and data_ingestion_interval into a single concept #444

Two variables represent the same semantic concept — "data resolution interval" — but are named differently depending on the component:

• prometheus_scrape_interval: used by asap-query-engine (both PromQL and SQL paths) for timestamp alignment and sketch lookup validation
• data_ingestion_interval: used by asap-planner-rs in SQL mode for generating streaming windows

The query engine has no notion of data_ingestion_interval; the planner has no notion of prometheus_scrape_interval in SQL mode. Callers (e.g. experiment_run_clickhouse.py) must know to pass data_ingestion_interval as prometheus_scrape_interval to the query engine — a sharp edge that caused the bug in #443.

Pick one name (likely data_ingestion_interval or a neutral resolution_interval) and thread it consistently through both components and their configs.