All GFS 0.25° grib files provided by NCEP/NOAA after this upgrade do contain announced changes. I don't know if Windy downloads grids of all available parameters/levels: data is huge (>0.5 GB for each forecast), while downloaded grids may be filtered by both coverage, forecast, parameter and level.
Obviously not all available parameters and levels are available via Windy UI.
Anyway, introducing all parameters and levels on UI would be useless for most of the Windy users (e.g. some parameters are available for levels 0.1 hPa, 0.2 hPa, etc.); they would be only useful for research purpose.

@petra-pik i tak uważam że jesteście najlepsi poporstu zauważyłem taki przypadek i informuje o tym ale bez tego da się żyć

Hi @udm196 right now the ECMWF is set as default. Thank you for your suggestion, we will consider it as a future feature.

@johnckealy
Weather forceast model is a fascinating topic ! The more you use them the better you understand them, you know when one model can be more accurate than another, what model can or cannot do. IFS is probably one of the best model to predict the changes at synoptic level (1000 or 2000km) when Arome is capable to solve and calculates convection which can be useful to predict thunderstorms.

Of course we can keep using one model only but Windy offers the possibility to easily compare the results from the different models so it is a nice and easy way to check the consistency between the models and therefore give you more confidence about the weather forecasts.

@WildOrangutan

+1
Other weather apps have a more practical design - allowing the user to preselect their preferred weather model - which is then the source for both app and widget.

Windy, sadly, uses ECMWF as the default which then, unless ECMWF is your preferred model, has to be changed every time the user opens the app and renders the widget useless.