It was only a matter of time before one of the operational runs positioned the cut-off low in a manner conducive to a major shot of heat (never mind the temps shown; we know GFS undercooks severely).

A thundery breakdown then occurs Friday 13th. Nearly all runs of the past couple of days from this model have gone with that, which is very annoying when you've got tickets to Goodwood Festival of Speed for that day! Standing out in fields with a vast amount of metal around you during thunderstorms... is not a good idea!

A return to settled weather by the following Sunday or Monday keeps being shown too, just to rub it in .

There seems to be a new split opening up between models (again, generally between GFS on the one hand and all the others). They all seem to be agreed on a bit of instability this week then ridging across the country in time for the weekend. But what happens thereafter is really diverging:

GFS has the ridge losing its link to the main AH and a cut-off low building to our South West. Pressure starts to drop as the low drags up impressively hot uppers into the UK. It then turns unsettled and westerly.

That is the "12-15th July 2003" solution.

ECM and GEM (too early for UKMO but seemed to be going same way) have the ridge keeping its link to the AH, the cut off low not really developing and the high slipping a bit to the WNW, eventually bringing the promise of cooler fresher air though keeping it dry.

That's more like the May/June 2018 solution.

I prefer the ECM/GEM version of events. Much less hot in the short-medium term but a more promising prognosis longer term. Above all we need to keep things dry - there's too much summer left to be getting into thundery breakdowns and wet spells.