It looks like there are multiple problems with the code in gridded mode that saves SOILWAT2 output. This is a general issue to house the issues until they are identified and documented at a finer scale. The problems:
The yearly output file not associated with soil layers is incorrect: the year column is screwed up and there are duplicates and some resulting strange values in at least the PRECIP column and probably more.
There are also redundant years in the monthly and daily output files.
There are no average output files across all cells.
It looks like there are multiple problems with the code in gridded mode that saves SOILWAT2 output. This is a general issue to house the issues until they are identified and documented at a finer scale. The problems: