Euro seasonal is better than the others but in an absolute sense, it is still pretty inaccurate. It absolutely shit the bed last year even on the October and November versions IIRC.
I also like to look at the H5 anomalies and not the 2m temp anomalies...they often don't seem in sync and the H5 anomaly forecast is going to be easier for the models to hit. IIRC, back in 2013-2014 and 2014-2015, it was showing these monster ridges over AK/EPO region (that largely verified) but they had warm anomalies in southern Canada and into most of the CONUS which is totally at odds with that pattern. So the H5anomaly was a lot more accurate for forecasting the sensible wx than the 2m temp anomaly.