After you have your shifts gruped correctly and ready to process - like in your previous question , you may use that as a subquery to get the type of the workshift.

SELECT *, CAST(TimeCheckIn as DATE) WorkshiftDay, 
      CASE 
         WHEN DayCheckIn<> pairs.DayCheckOut AND  pairs.HourCheckOut < 11 THEN 'S3'
         WHEN HourCheckIn BETWEEN 5 AND 7 AND HourCheckOut BETWEEN 13 AND 18 THEN 'S1'
         WHEN HourCheckIn BETWEEN 10 AND 15 AND HourCheckOut BETWEEN 21 AND 23 THEN 'S2'
       END AS 'WORKINGSHIFT'
FROM 
(
    SELECT IDUser,        
           MIN(TimeCheck) AS TimeCheckIn,
           MAX(TimeCheck) AS TimeCheckOut,
           DATEPART(hour,MIN(TimeCheck)) AS HourCheckIn,
           DATEPART(hour,MAX(TimeCheck)) AS HourCheckOut,
           DATEPART(day,MIN(TimeCheck)) AS DayCheckIn,
           DATEPART(day,MAX(TimeCheck)) AS DayCheckOut
    FROM (
      SELECT *,
             ROW_NUMBER() OVER (PARTITION BY IDUser ORDER BY TimeCheck) AS rn
      FROM TBLLOGFULL
    ) AS t
    GROUP BY IDUser,  FLOOR((rn - 1) / 2)
    HAVING COUNT(*) = 2    
) pairs
ORDER BY IDUser

https://dbfiddle.uk/0Q8hKRFm

I.e. this is the second part of the problem, which can be written like:

SELECT *, CAST(TimeCheckIn as DATE) WorkshiftDay, 
      CASE 
         WHEN DayCheckIn <> DayCheckOut AND HourCheckOut < 11 THEN 'S3'
         WHEN DayCheckIn = DayCheckOut  AND HourCheckIn BETWEEN 5 AND 7 AND HourCheckOut BETWEEN 13 AND 18 THEN 'S1'
         WHEN DayCheckIn = DayCheckOut  AND HourCheckIn BETWEEN 10 AND 15 AND HourCheckOut BETWEEN 21 AND 23 THEN 'S2'
       END AS 'WORKINGSHIFT'
FROM WorkShiftsView 
ORDER BY IDUser

and

CREATE VIEW WorkShiftsView AS
    SELECT IDUser,        
           MIN(TimeCheck) AS 'TimeCheckIn',
           MAX(TimeCheck) AS 'TimeCheckOut',
           DATEPART(hour,MIN(TimeCheck)) AS HourCheckIn,
           DATEPART(hour,MAX(TimeCheck)) AS HourCheckOut,
           DATEPART(day,MIN(TimeCheck)) AS DayCheckIn,
           DATEPART(day,MAX(TimeCheck)) AS DayCheckOut
    FROM (
      SELECT *,
             ROW_NUMBER() OVER (PARTITION BY IDUser ORDER BY TimeCheck) AS rn
      FROM TBLLOGFULL
    ) AS t
    GROUP BY IDUser,  FLOOR((rn - 1) / 2)
    HAVING COUNT(*) = 2

Thus you can separate the previous logic of finding shifts and solve the two problems separately. You see that then the second task is very simple, the tricky part is to get your shifts data from chronological log.

https://dbfiddle.uk/X6rfaZyv

If you want to solve another problem of duplicate checks, you should add another layer to remove duplicates (less than 5 minutes) like the code bellow and use FixDuplicateChecks view instead of reference to TBLLOGFULL table:

https://dbfiddle.uk/YvjH14Ro

CREATE VIEW FixDuplicateChecks AS
SELECT T1.IDUser, T1.TimeCheck
FROM TBLLOGFULL T1
WHERE NOT EXISTS (
   SELECT 1
   FROM TBLLOGFULL T2
   WHERE T2.IDUser = T1.IDUser
   AND T2.TimeCheck > DATEADD(MINUTE, -5, T1.TimeCheck)
   AND T2.TimeCheck < T1.TimeCheck
)