I have now finally gotten time to sit down and try answer this question (in the blog title) in great detail because it involves a lot of explaining to do and complications. I have personally not witnessed a moment like this in my model watching years so far!
Firstly, yes it's multiple factors that are possibly causing this high uncertainty we've been having during the latter part of the Summer after the Atlantic broke through on the 27th July. Before this period, the Polar Vortex was weak and so was the jet stream that throws the Atlantic depressions at us normally. This allowed there to be reasonably high certainty for a long time through Spring into early Summer as there wasn't much to really break down the big anticyclone and all the dry weather even systems like Hector in mid-June failed to do so showing you just how stubborn the conjoined Scandi High and Azores High were which is evident in many of our hot summers. However, as shown by the graph below of zonal wind speeds, the Polar Vortex has gathered pace as it usually begins to do at this time of year anyway. Another view of this is looking at the stratosphere temperatures - you'll find the chart below. The grey line represents the 1981-2010 average stratosphere temperatures at 10hPa whilst the black line is the actual recorded temperatures through the year. As you can see by both lines, they're on their way down. A colder stratosphere equates to an intensified jet stream whilst a warmer stratosphere equates to the opposite with a meandered or weak jet stream (usually involved with sudden stratospheric warming events). This is completely normal as I said and that it happens every year from August into early Autumn whilst numerical weather predictions try to get a grip of how the Polar Vortex is going to behave. However, the question lies in why is it even more difficult (and exceptionally so) this time around even if it is normal. Well I'm using this part as a starting point to the post because whilst it's normal, it plays an important part. Basically, it's a nuisance than anything else.
I previously gave a post about the rossby waves which you can find here on the weather forum of Boards.ie: https://www.boards.ie/vbulletin/showpost.php?p=107680385&postcount=382
As the rossby wave occurred, it gave the jet stream a bit of oomph and since, we have been in this changeable pattern to which Ireland is known for. I expected this when I theorised about it. These rossby waves have slow motion and as a result, we could be in this changeable pattern for a good while unless something drastic happens like a big Atlantic tropical cyclone ascending high pressure from the Azores (or the back-to-school week phenomena occurs). The changeable pattern is going to bring uncertainty into the question at times because well..... it's in the name, changes happen regularly with interchangeable air masses, in contrast to a blocking pattern like we had from May to early July which has not much going on.
Since the SSW event back in February, the North Atlantic SST profile has been in a position not seen since at least the early 90s. This is the last time when there was a cold Atlantic Multidecadal Oscillation (AMO). The AMO is a climate cycle that affects the sea surface temperatures of the North Atlantic Ocean. When the majority of the North Atlantic is cold, the index tends to be negative whilst the reverse occurs when the majority of the North Atlantic is warm. It is prone to changes on a regular basis like the other indexes but it comes in cycles as an average overall. For example, negative AMO occurred from the early 1960s to the early 1990s but since the early 2000s up to at least 2017, the AMO has been positive. A sign of negative or cold AMO is a horseshoe shape of cold sea surface temperature anomalies as shown below in May 2018. There was theories going around, including by the UK Met Office and myself, that this meant we'd get a hot and dry Summer with a lot of anticyclonic influences as it shared similarities to how May 1976 looked. Well, that theory turned out quite well. However, this may have influenced the uncertainty because of observations struggling to understand the interaction of this with the Polar Vortex gathering pace and not to mention how we have been in the positive AMO phase since the mid-90s. You can see the latest and this time last year's Atlantic SST profile charts below the May 2018 one. They are night and day different in many ways including a warmer tropical Atlantic (which led to an active hurricane season combined with the La Nina event) and to the south of Greenland. This cold AMO is very unusual for modern times and thus, the models don't really have a lot of data for this. It will be interesting to observe. Due to how 1976 went with an exceptionally wet Autumn following the hot and dry Summer, you'd think an unsettled Autumn is the likely scenario especially with the Scandi High broken down which was one of the main elements in developing this hot Summer - similar in 1976.
We would usually be seeing uncertainty caused by Atlantic tropical cyclones and hurricanes at this time of year up to October but this year, there is a possible El Niño forming and there is also the cold sea surface temperatures over the main development region (MDR) for hurricanes in the North Atlantic. In El Niño years, vertical wind shear is increased over the Caribbean and Atlantic and this helps to prevent tropical cyclones forming into hurricanes. The opposite occurs over the Pacific. This is what has happened this year, the vertical wind shear has been above average and it has been a quiet season so far with agencies expecting this to continue through the Autumn possibly making 2018 the quietest Atlantic hurricane season since 2015 at least. This should make the weather straightforward for us but of course it doesn't, mother nature is never that nice to us! ENSO is also proving to be a lot of headaches itself, see the latest ENSO blog post for information on the matters here: https://bruener45.wixsite.com/britishislesclimate/blog-1/enso-update-13-august-2018
One common thing that every model watcher knows or should know is that the models do not handle pattern changes well but we've been in this changeable pattern for more than a week now, you would think they'd gain some certainty.
There is a lot of possibilities that are causing the uncertainty to be abnormally high. Let's compare model runs.
This is every GFS run for Friday 24th August during the afternoon since Sunday just gone. Very different from run to run, day to day.
Meanwhile, the ECM has been ok since the 24 August came into its timeframe but look how different today's 12z was to say yesterday's 0z:
It's all really complex and probably a bit too hard to understand but I've tried explaining it as simply as I possibly can without getting a bit "too sciencey". It's only going to get worse into Winter if all these weird phenomena occur. This is not just uncertain because we've been used to high certainty and consistency in model outlooks during the Spring and early Summer but also down to reasons named above with millions of further potential culprits waiting to be seeked. It's a a matter of wait and see. I hope this post has answered your question.