Actually you could be causing a performance problem by getting SQL to use the same query plan for different values, i.e., by avoiding the "overhead" of a recompile.  The first...

Below is some code that will probably do what you want.  I strongly recommend you run it first with the EXEC() commented out.  If all of the output looks ok,...

If I understand correctly, you just need to add the condition on the JOIN:

...previous_part_of_query_same_as_before...

INNER JOIN #temp_Lookback t2 ON t1.Id = t2.PreviouslyCompletedId AND t2.Type = 'F'

You'd typically want to use dynamic SQL for this, so that only the relevant columns were included in the comparisons.  You'll also typically want to recompile to get a specific...

It definitely could, esp. if there's an index containing the type column.

First try refreshing the view.  The "*" in the view may be causing some issues if the table has changed after the view was created.

I have to disagree.  If you use...

I'd say it depends on how often you need to show the resolution duration.  That's a rather complex calc to do repeatedly.   And presumably the Resolution Time would only be...

Sorry.

SET ANSI_NULLS ON
SET QUOTED_IDENTIFIER ON
GO
CREATE FUNCTION [dbo].[DelimitedSplit8K] (
    @pString varchar(8000),
    @pDelimiter char(1)
    )
RETURNS TABLE WITH SCHEMABINDING 
AS
/*SELECT *...

A few other things need changed to match SQL Server syntax as well.

CREATE TABLE "circuits" (
  "circuitId" int NOT NULL IDENTITY(1, 1), --<<--
  "circuitRef" varchar(255)...

Let's fall back on the old reliable dbo.DelimitedSplit8K function, to avoid doing all the parsing ourselves.

;WITH sample_data AS (
    SELECT * FROM ( VALUES
...

UPDATE ...

SET FirstName = PARSENAME(LEFT(EmailAddress, CHARINDEX('@', EmailAddress) - 1), 1),
    LastName = PARSENAME(LEFT(EmailAddress, CHARINDEX('@', EmailAddress) - 1), 2)

Yes to the most RAM you can get and actually use for SQL Server: 128GB for Standard Edition, as much as you can afford for Enterprise.

Note that, if Standard Ed,...

You can write them to a table -- add an identity for ordering -- and then SELECT from that table, for example:

SELECT sql_text AS [--sql_to_run]

FROM #some_table_name

ORDER BY $IDENTITY

You didn't say what the result should be if it is tied, for example, 2F and 2M?  In that case, I went with 'FemaleMix', adjust that if you need to.