If you are using Section Access to hide data sets for individual users (by way of Data Reduction) then set analysis cannot undo this because the data isn't there anymore.
I guess you are talking about showing benchmarking data together with filtered individual performance data tailored to individual users? You can pre-calculate these benchmark data sets (aggregated so that all identifying characteristics are removed) and associate them with all individual users. That way, every user will keep individual data and benchmark data available after data reduction has removed everything else.
