Hmm, you are grouping by the same columns you are averaging, so that probably isn't giving you what you want. 
Also the OVER clause is only being applied to the...

Hmm, it shouldn't.  I suspect the error is actually somewhere else.  
The aliases I put on the average column results are invalid because they start with a number.  Try removing...

Ah, so for every row, you want to average each of the SSI columns for all rows within 24 hours prior to the current row date/time?

I doubt this...

Sorry I didn't figure you needed the end date, since you have it as the start date of the next row. 
If you really need that on the same...

What are you stuck on?

Would something like this work for you?

SELECT
    Location,
    DateValue, Sum(Number)
FROM
    #_tempData TheData
JOIN
    (
        SELECT     StartDate AS DateValue
        FROM...

Here's an example using just table joins:

;WITH TabData (Code, Date, Type)
AS
(
    SELECT
        Code, Date, type
    FROM
        @tabData
)
SELECT
    Full_Data.Code,

Should the SELECT have a try_cast in it as well?
My understanding is that with larger data sets, it is possible for the optimizer to validate results before applying filters.

Three or those are included due to commas in the name, which have not been excluded by the pattern.  I think you are past the problem of special characters in...

Interesting... it looks like escaping the square brackets only works when you explicitly specify the escape character for some reason?

Give this a shot:

DECLARE...

The underscore is a single character wildcard in LIKE expressions.
You'll need to encapsulate it in square brackets:

Select * 
from [XLSFiles] 
where fullimagepath like '%[^a-zA-Z0-9.&[_]-#$\\()...

Add another variable to hold the previous group value, so you can store the current and previous values in variables.
On each loop, update current value via cursor, compare it...