It's fascinating (from a nerdy point of view) just how poorly the ensembles are coping with the potential very hot conditions on Thursday. ECM, for example, shows this IMBY:
The maximum from the 51 members is 31.1C, but the op has 35.6C - not just slightly warmer, but a massive outlier. However, given the noises from the Met Office (their text forecast goes for 33C in the SE tomorrow as a high, for example), it's clearly not thought unusual.
The GEFS also has an "outside the pack" forecast, albeit to a lesser magnitude than ECM.
Generally the ensembles offer excellent guidance, but it seems their resolution isn't as high as it could be, hence small pockets of extremely high (or low) temperatures can get missed.