G, if you look at the extraction code described in post #1 and in particular look carefully at the starting position and size of field N, then the starting position of field N+1, there is no apparent delimiter so STRING_SPLIT() might not work here.
I only use STRING_SPLIT to divide the lines into separate records because the original records I work with have a whole text file in a single record. OP doesn't require this step because the records are already separated in the source recordset.
Clearly it wouldn't be useful for splitting fixed width data with no delimiters.
BTW One must take care with SPLIT_STRING because, despite being numbered, the order of the output is not guaranteed to match the input. In my case I am not concerned about the order of the lines so it is a perfect solution.