I don't think "AVG" really has any meaning for the diff between only 2 dates?!

SELECT cust_id, DATEDiff(day,Min(visit_date),Max(visit_date))
FROM (
  SELECT cust_id, visit_date
  , ROW_NUMBER()...

Try forcing a MERGE join, see if that helps significantly.  If it does, then go ahead and cluster the SOH table on (  IT_ID, ST_ID, DateID ).  Yes, that means the rows...

Partition the clus columnstore index on that date.  Then SQL can eliminate partitions with non-matching dates.  I partition by month on many of our very large tables, but you might...

It's definitely possible to do.  You'll most likely need an index on the table to support the lookup.  You might be able to use a filtered index, but I can't...

SELECT ut.IdUser,
    MIN(ut.IdShop) AS IdShop_Min, MAX(ut.IdShop) AS Id_Shop_Max /*MIN & MAX are optional, of course*/
FROM @UserTransaction ut
GROUP BY ut.IdUser
HAVING COUNT(DISTINCT ut.IdShop) >...

Still seems too long for only 76K rows.

1) Check fragmentation on the clus index. If it's bad, the index may need reorganized or rebuilt.
2) Compress the table...

SELECT GL_TransLines.[GLL_Account] as "Account"
    , GL_TransLines.[Amount] as "Current Expenses"
    , GL_TransLines.[COP_Period] as "Period"
    , GL_TransLines.[GLT_TransDate] as "Trans Date"
    ,...

Actually that's a significant understatement by me.  Your clus index is generally superb (although you could probably remove the Attribute* columns with no real...

Your clustered index is fine.
You really don't need the temp tables for this specific query, but if building them is not taking much time, then that's OK too.

For testing a datetime/datetime2 column for the current day, you should do this:

--good way
cr.CreatedDate >= DATEADD(DAY, DATEDIFF(DAY, 0, GETDATE()), 0) AND 
cr.CreatedDate < DATEADD(DAY, DATEDIFF(DAY, 0, GETDATE()) +...

Data warehouses are often denormalized.  However, this data is so unlikely to be needed that all columns other than, say, FullName, should be put into a separate table, which can be joined...

Back in the day, I would do a pattern search to determine eligible data types:

SUM(CASE WHEN column NOT LIKE '%[^0-9]%' THEN 1 ELSE 0 END) AS column_has_only_digits,
etc.

For a first pass, you could try just ISDATE, IS NULL, ISNUMERIC and a few other key attributes:

CREATE TABLE dbo.column_profile ( ... );

INSERT...

Put the metrics in a table.  Then we can join to each metric as needed.