When using Visual Analytics (version 7.2) for external reports, I very much recommend a functionality to hide og mask specific cells.
Working with detailed data and plotting cross-tables on multiple categoric variables, you will often run in to combinations and statistics based on a limited number of observations. As such statistics can very well be insignificant, it should be possible to hide these and set up specific criterias for hiding - that is, without taking out any data.
In VA it is possible to filter out fx rows in a table based on a specialized filter. But this leads to taking out the specific observations, and it changes all aggregate sums in the table.
In VA it is also possible to change the style based on specific expressions. Fx, I can change the color of the cell-values in a cross-table to red, indicating an insignificant value. The expression could be a criteria on the no. of observations that the value is based on ie. the frequency. Unfortunately this criteria can only be defined if the frequency column is already in the the table - and I'm not interested in showing no. of observations in an external report.
A way out (and maybe an easy one to start with) could be making it possible to change style based on expressions on all possible dependent variables in the data.
In my opinion, masking and hiding data to external recipients should be regarded as very important if VA is considered a tool for external reporting. If you cannot control data completely all value of making external reports disappears.
For now, I will have to aggregate data in SAS BASE defining my criterias in data before loading them into SAS VA. This closes down all analytic possibilities in VA, and I might as well just make an Excel report for my clients instead.