GEFS 06z does seem to have firmed up on longer term cold. On the subject of ensemble "blips" I started to notice an increased frequency of them after the SSW last winter. The same thing happened several times during the summer when 850s bulged upwards at relatively short notice. Obviously at longer range the scatter means colder and milder outcomes cancel each other out to a degree. However, I think there is more than that going on. An increased incidence of high pressure blocking is the key I think, because it means the potential for significant deviations from the norm are often only a few hundred miles away.
I tend to agree. Often you can find a very different ensemble profile if you select a location even a hundred miles away - particularly if you choose the locations carefully having regard to the prevailing synoptic pattern.
The ECM 00z ensemble suite for London carries an identical message to the GEFS, i.e. long-term signal for cold weather beyond a very brief blip next Saturday. Most of the variants bring in the warm sector at some point that day but a reasonable cluster delays it until the evening. It's noticeable that there's a near-total absence of mild outcomes and very few offering normal temperatures.
Of course we know that ensemble suite can also flip but as it stands the outlook remains a cold to very cold one - and I expect the Met Office forecast to remain largely unaltered.
Location: South Cambridgeshire
130 metres ASL
52.0N 0.1E