Esta lógica de verificação do índice está correta?

Estou escrevendo um script para verificar rapidamente os índices em um determinado banco de dados. Acho que tenho o script escrito ok, mas para remover linhas duplicadas, fiz várias SUMfunções e não tenho certeza se isso está correto ou não:

SELECT      OBJECT_NAME(i.object_id)                                                                                                                AS 'Table Name'
,           COALESCE(i.name, '')                                                                                                                    AS 'Index Name'
,           i.type_desc                                                                                                                             AS 'Index Type'
,           i.is_primary_key                                                                                                                        AS 'Primary Key'
,           i.is_disabled                                                                                                                           AS 'Index Disabled'
,           ROUND(CONVERT(FLOAT,SUM(user_seeks)) / (CONVERT(FLOAT,ISNULL(NULLIF(SUM(user_scans),0),1))),2)                                          AS 'Seeks Per Scan'
,           ROUND((CONVERT(FLOAT,SUM(user_seeks) +SUM(user_scans) + SUM(user_lookups))) / (CONVERT(FLOAT,ISNULL(NULLIF(SUM(user_updates),0),1))),2) AS 'Reads Per Write'
,           (SUM(user_seeks) + SUM(user_scans) + SUM(user_lookups))                                                                                 AS 'Reads'
,           SUM(user_updates)                                                                                                                       AS 'Writes'
,           SUM(user_seeks)                                                                                                                         AS 'Seeks'
,           SUM(user_scans)                                                                                                                         AS 'Scan'
,           ROUND(AVG(ps.avg_fragmentation_in_percent),2)                                                                                           AS 'Avg Fragmentation %'
,           NULLIF((SUM(leaf_insert_count) + SUM(leaf_update_count) + SUM(leaf_delete_count)),0) / NULLIF(SUM(leaf_allocation_count),0)             AS 'Leaf Page Splits Per Write'
,           NULLIF((SUM(nonleaf_insert_count) + SUM(nonleaf_update_count) + SUM(nonleaf_delete_count)),0) / NULLIF(SUM(nonleaf_allocation_count),0) AS 'Non-Leaf Page Splits Per Write'
,           CAST (100.0 * SUM(row_lock_wait_count) / (1 + SUM(row_lock_count)) AS NUMERIC(15,2))                                                    AS 'Block %'
,           SUM(row_lock_wait_in_ms)                                                                                                                AS 'Row Lock Wait (ms)'
,           CAST(1.0 * SUM(row_lock_wait_in_ms) / (1 + SUM(row_lock_wait_count)) AS NUMERIC(15,2))                                                  AS 'Average Row Lock Wait'
FROM sys.indexes i
LEFT JOIN sys.objects o ON o.object_id = i.object_id 
LEFT JOIN sys.dm_db_index_operational_stats(NULL, NULL, NULL, NULL) s ON o.object_id = s.object_id AND i.index_id = s.index_id
LEFT JOIN sys.dm_db_index_usage_stats u ON s.object_id = u.object_id AND u.index_id = s.index_id
LEFT JOIN sys.dm_db_index_physical_stats (DB_ID(), NULL, NULL, NULL, NULL) AS ps ON u.index_id = ps.index_id AND u.object_id = ps.object_id
WHERE OBJECTPROPERTY(o.object_id,'IsUserTable') = 1 
AND NOT OBJECT_NAME(i.object_id) = 'sysdiagrams' -- exclude sysdiagrams
--AND i.is_primary_key = 0
GROUP BY OBJECT_NAME(i.object_id), COALESCE(i.name, ''),i.type_desc,i.is_primary_key,i.is_disabled
ORDER BY OBJECT_NAME(i.object_id), COALESCE(i.name, '')

tudo bem? Ou SUMas linhas estão ruins?

Além disso, há alguma informação que você acha que seria útil para encontrar índices que não são usados/precisam de reindexação, etc.