When analyzing skewed data, median can be a better measure than mean.
Currently this can be implemented with a User Defined Function (e.g. http://www.mooreds.com/wordpress/archives/376), however this requires more overhead when upgrading and deploying database. This can also be handled with a query (e.g. http://stackoverflow.com/questions/1291152/simple-way-to-calculate-median-with-mysql), however that has a performance hit.
Would adding a new aggregate function be possible?