The 2 decimal "output accuracy" of r-expectancy score itself is not the problem. Something within the formula is being rounded way too much. You can see the "gan line pattern" in the r-expectancy chart, (everything is rounded to the nearest line) the more trades there are the less accurate (more rounded) r-expectancy SCORE becomes, also, with 1k trades in OOS, you can see in the table many 0's and nothing in between ~ 2.7 and ~ -2.7.
EDIT: The solution is to base the r-expectancy SCORE column on a more accurate r-expectancy column. By simply creating an r-expectancy column that is accurate to 4 decimals instead of two and then creating an r-expectancy SCORE column based on this more accurate r-expectancy column I was able to eliminate the rounding issues. To be clear: R-expectancy rounded to 4d, R-expectancy SCORE based on the more accurate R-expectancy but R-expectancy SCORE still rounded to 2d
The solution is to base the r-expectancy SCORE column on a more accurate r-expectancy column. By simply creating an r-expectancy column that is accurate to 4 decimals instead of two and then creating an r-expectancy SCORE column based on this more accurate r-expectancy column I was able to eliminate the rounding issues.