A few critical tables would be a good start, particularly large child tables that are currently clustered on an identity column.

Edit: Be sure to include any "missing index"...

DECLARE @Begin_Period int;
DECLARE @Last_Period int;

SET @Begin_Period = 201712;
SET @Last_Period = CASE WHEN @Begin_Period % 100 < 12 THEN @Begin_Period + 1 ELSE @Begin_Period...

The code below will produce two result sets.  The first is missing index stats/info.  The second is existing index stats/info.  We combine the two -- and often other needed...

Another option could be to use a static view that is dynamically generated for the formulas.  When a formula(s) change, the view is regenerated.  Just an idea.

The first focus should be on getting the best clustered index on all (significant) tables.  That is the single most critical performance factor for most tables.  Hint: This is most...

Looks rather straightforward, since the detail table already contains the Position the person was in on that given day.

If  you could post useable sample data -- CREATE TABLE...

The trigger would still fire, but you could have it immediately exit.  The easiest way is likely using CONTEXT_INFO.  Have the app set the first two bytes of it to...

Common, but not the best.  You'll hurt your overall performance significantly this way.

These...

You don't need or want the dashes in the date:

SELECT STUFF(STUFF(STUFF(main_datetime, 13, 0, ':'), 11, 0, ':'), 9, 0, ' ') + '.0000000' AS datetime2_format,