It seems that the coefficient 0.278 is not used in the demo calulation, as the result of 8.605184 is not correct.
The actual code is probably correct because the coefficient appears in the calculation of Rad1wh, but you know the code better to decide if any kind of correction needs to be made.
https://github.com/kilianknoll/DWDForecast/blob/645e0ade556083b4fbb57bba8733f881a8896259/configuration.ini#L77
It seems that the coefficient 0.278 is not used in the demo calulation, as the result of 8.605184 is not correct. The actual code is probably correct because the coefficient appears in the calculation of Rad1wh, but you know the code better to decide if any kind of correction needs to be made.
Thanks by the way for this insightful approach!