Currently the OVER function is, frankly pretty poorly implemented for anyone wanting to do data analysis that's not time series data.
OVER functions should be able to access aggregations in parenthesis. For example, if I wanted to calculate the average for an entire dataset, I would do something like
Avg([Reference Voltage]) OVER (Intersect(All([Axis.Rows]),All([Axis.Columns]))) as [Data Average]
However, suppose I also wanted to calculate the average+6 sigma from that. I am not allowed to do
( Avg([Reference Voltage])+6*StdDev([Reference Voltage]) ) OVER (Intersect(All([Axis.Rows]),All([Axis.Columns]))) as [6σ Limit]
There are a LOT of pain points when using custom expressions to aggregate data and they really need a rework.