Version 1.8.4 is online
This is maintenance and includes a - I think important, at least for Harm who uses m/s for wind speed - fix for the unit issue. Until now prediction could only be used if the units of the user were km/h, deg C, mm and hPa. Now all possible settings possible of units in Cumulus should work equally well with the pwsFWI prediction feature. If you use deviations of units, please look sharply for anomalies (aka bugs). Still waiting for my first imperial unit user.
I am looking forward to your analysis sheets, If you have 7 predictions with corresponding calculations, that should be enough I guess. (meteosangonera's sheet can be followed online, PaulMy's sheet you can read yourself). The first indications are surprisingly - yes, I am surprised

- accurate predictions if we use the two day ahead prediction.