This might be obvious, but have you checked that it's not blocked by an open transaction?
It happened to me once.

And yet another option:

WITH CTE AS(
  SELECT *,
   LAG(cs.Value1) OVER(PARTITION BY cs.DuplicatingValue ORDER BY cs.Value1 ) prev_Value1
  FROM dbo.CentralSample AS cs

Here's a different option that might be better, but you should still test for performance.

WITH CTE AS(
  SELECT *,
   ROW_NUMBER() OVER(PARTITION BY...

I started interacting with Pervasive SQL and it's frustrating. Why do people pay for a product that won't work as well as many other free products that have been well...

Maybe this would make the code easier to read and prevent any SQL injection.

DECLARE @sql NVARCHAR(MAX);

SELECT @sql = N'SELECT DISTINCT
  PARSENAME(SchemaName, 1)...

For more information on...

Something tells me that you would benefit more by having the clustered index on StartDateTime. I don't see how or why ExecutionID can be relevant to your queries.

Why would you create a table, insert rows and then select the new table?
Why can't you simply do a straight select?