How to handle DISTINCT COUNT against multiple columns???

Here is my problem:  The business requires a measure to be created that involves identifying the distinct count of a field for various columns in specific dimensions.  The field is a store ID (Retailer_ID).  I need to capture the distinct number of store id's across the Distributor Dimension..which contains the following fields:

• Business Unit
• GM
• MDM

Also this will be at the monthly grain.

So here is the SQL that I have:

SELECT

        bu_cd, bu_dsc, gm_cd,  gm_dsc, mdm_cd,  mdm_dsc,

    count(distinct case when date_id between 20130101 and 20131231 then f.retail_id else NULL END) acct_sold_mth_ty,

        count(distinct case when date_id between 20120101 and 20121231 then f.retail_id else NULL END) acct_sold_mth_ly,

        count(distinct case when date_id between 20110101 and 20111231 then f.retail_id else NULL END) acct_sold_mth_lly     

FROM

    datamart03.dbo.tbl_asa_fact f

    join datamart03.dbo.tbl_asa_dist d on d.dist_id = f.dist_id

GROUP BY


bu_cd, bu_dsc, gm_cd,  gm_dsc, mdm_cd,  mdm_dsc,

ORDER BY

        1 ASC, 3 ASC

I understand that the distinct count is a killer when trying to aggregate against large amounts of data, especially like how I have it.  

The various permutations that are required in order to get the accurate measure, is where I am severely challenged....in fact what I have shown above is a small sample of what needs to be done.  There is in fact another dimension with 3 columns that would also need to be part of the group by!

 I am hoping that possibly there is someone out there with a more elegant way of handling this??

Thanks in advance!!