I am so sorry it has taken me so long to get back here to further comment, however, the widget has not had so much as a hiccup since my last install in May. I do want to thank everybody who tried to help me. And I still love Windy and I still miss having a radar widget, so if someone finds an app that has a radar widget for iOS, please pass the information my way. Please.
And again, thank you, all.
@petra-pik Many thanks, I've already used the e-mail link. Actually I was not complaining about e-mails I received.
I just tried to emphasize cases were, in current implementation, removing Alerts could be fastidious in practice.
As I saw there are other posts on same subject, I think all these cases could be resolved if new, dedicated Alerts Management would be available on UI.
Sometime there is a quite bad synchronisation between the basic Meteogram and the map for a favorite location. It seems that this is linked to the way how you move the time slider. I recommend to use the L&R arrow keys of your keyboard to move the time slider, it allows to do it more precisely hour per hour. The synchronisation seems better, but you still can get 1 km/h max discrepancy with Wind gusts.
@jonwilloughby Hi, you are right, see: https://apps.ecmwf.int/codes/grib/param-db?id=228057
But to be honest, these numbers probably do not fit with real lightning count in the specific storm.
According to https://www.ecmwf.int/en/newsletter/155/meteorology/promising-results-lightning-predictions , part "Lightning parametrization", the result is normalized by constant α "where α is a tunable coefficient, currently set to 37.5 to match the annual global mean flash rate". So it is expectable that forecasted annual lightning count may be correct, but in the specific storm? who knows...
Nevertheless the Thunderstorm forecast is not intended for calculating flashes, but to predict locality, probability and severity of storms. And as you can see on case of Crew Dragon launch, the boys from ECMWF predicted that storm very well :-)