TIMESTAMP_NTZ is scanned as a zoned TIMESTAMP #506

TIMESTAMP and TIMESTAMP_NTZ both come back from rows.Scan as time.Time, and ColumnTypeDatabaseTypeName reports TIMESTAMP for both. A zone-less value cannot be told apart from an instant, and it is passed through a location conversion.

On the Arrow batch the two types are different. Both are timestamp[us]. The TIMESTAMP field has a time zone set. The TIMESTAMP_NTZ field has an empty TimeZone. Nothing in the driver reads that. GetDBTypeName returns the Thrift primitive name with _TYPE stripped, and there is no TIMESTAMP_NTZ handling in the driver. Every *arrow.TimestampType uses one converter, built from arrow.FixedWidthTypes.Timestamp_us.
Reproducer, against a SQL warehouse:

SELECT
  TIMESTAMP '2024-06-15 09:15:00.123456' AS zoned,
  TIMESTAMP_NTZ '2024-06-15 09:15:00.123456' AS ntz

var zoned, ntz time.Time
rows.Next()
rows.Scan(&zoned, &ntz)
// both are time.Time
// ColumnTypeDatabaseTypeName(0) == "TIMESTAMP"
// ColumnTypeDatabaseTypeName(1) == "TIMESTAMP"

Expected: TIMESTAMP_NTZ keeps the wall clock it was given and is reported as TIMESTAMP_NTZ. TIMESTAMP stays an instant https://docs.databricks.com/aws/en/sql/language-manual/data-types/timestamp-ntz-type.