This should greatly simplify the build of a query and reduce the need for a secondary data structure such as a pivot table.

The performance of query executions would be greatly enhanced by allowing the DB to use native aggregation functions, thereby reducing the need for indexes to be created to handle filters required for multiple dimension filters.

Less data would need to be transferred over the WAN / LAN (this is important in scenarios where the Jet DW is housed remotely - eg. in the cloud) as it would already be aggregated - as opposed to returning all records and then pivotting and summing in excel.