ProbBins for forecast case
I would like to have time to think about this carefully since I am not sure if we need to modify ProbBins or to create a separate development. You can take a look at the code here: https://earth.bsc.es/gitlab/external/cstools/-/blob/master/inst/doc/UseCase1_WindEvent_March2018.R
From line 79 to line 95, it is computing a kind of ProbBins: it uses the reference data to calculate the thresholds and looks to which bin belongs each value of the forecast.
I don't consider this development urgent because I don't know other people using this. However, @lpalma may know if it has done a similar development for the operational.