I can remember last summer the GFS throwing bonkers temperatures of +43 out 10 days in advance, and had a good laugh. Again, in December they picked up the low that dumped 4-6 inches over the South East way in advance. Both were really specific and easily dismissed so far out, but despite their disastrous history at times, particularly January, they have a knack of identifying some these outlier events in some of their runs.
I lived in Barrie Ontario for 7 years where were got 2.3m of snow a year, mostly lake effect, and the same Euro vs GFS arguments raged there, but again I remember the GFS picking an 18 inch storm 10 days out, and the other models only caving in the last few days. The GFS won.
My concern about this cold spell, model variance aside, is that if things do align the height of the sun will start to be an issue especially as we approach mid March, but if the winds are strong, then that would remove the warming effect from convection. Again, where we lived in Canada was on the same latitude as Nice, so we could, and did, get +25 occasionally in mid March, but there were more years when the maximum temperature didn’t get above -10 in the day even in late March, especially if there was snow cover.
There are plenty of examples of serious cold in March in the UK e.g. 2013 (2018 obviously) I am getting hopeful of something, even down here in Sussex.
March 2013 United Kingdom winter storm - Wikipedia
EN.WIKIPEDIA.ORG