Whilst I agree generally, I can't help thinking that sometimes the UK is badly prepared for conditions that don't really deviate that much from the middle tercile. No-one is expecting perfect operation in the 1 in 200 year event such as the October 1987 storm or winter 1962/3, but major struggling in temperatures of 25-30C which occur somewhere in the UK virtually every year is poor in my opinion. Yes I would expect chaos if 1.5 meters of snow fell (which, since none of the national rail network goes over the Cairngorm plateau isn't going to happen) but I would not expect significant disruption if half an inch fell overnight (and I don't think it does to be fair). In other words the rail network ought to be able to cope with conditions that occur in more years than not, if not then it is too flaky and lacking in robustness.
From an engineering standpoint, I don't disagree at all with your words, but the railways, at least in the UK, are also a political football. There is a constant 'value for money' dialogue that applies constant pressure on both NR and TOCs to keep costs below some idealised notion of how much it should cost. This pressure is augmented by regular sniping by the largely right-wing press that grudgingly accepts its very existence because a significant proportion of their readerships rely on rail travel in their commutes or business requirements.
In my lifetime, the nationalised and more recently privatised network has been forced down money saving engineering paths that have ended in tragedy, e.g. unrealistic expectations of staff deemed to have caused the Clapham disaster and Railtrack's maintenance lapses that caused the derailment at Hatfield to mention just two.
Fortunately with a more safety aware culture generally, Network Rail's record on track failure is much better, the emphasis being on engineering a PW for reliability over a temperature range. As Bald_Rick says, reliability does fall off above about 23degC but fortunately, regular temperatures above that are still a minor part of our weather cycle. It is possible to design track to cope with temperatures over 40C, but that would be more expensive to install and maintain particularly with the density of traffic on major routes. The additional expense would trigger a chorus of criticism from those in the media who have nothing better to offer.
As far as the TOCs are concerned, they are hemmed-in by the one-dimensional culture of lowest bid wins. The DfT don't have to face the music when trains fail to cope with higher temperatures, but they do get dictats from the exchequer and No. 10 about how much they should allow. These days, virtually all new and recent trains in the UK are version of globally deployed deisgns. There are Desiros and Aventras operating in far more harsh environments that we ever get here. The Hitachi A100/200/300 trains on order will be version of designs that are running quite happily in both hotter and colder climates. What is different, apart from the slightly smaller loading gauge is the pressure to force prices down, leading to cut corners both at acquisition and during support phases.
I also fail to see what the mortality of vulnerable people has to do with the robustness of the rail network.
The mortality of vulnerable people is (or should be) a primary concern of all. For their part, the modern railway has a very good human safety record, and despite the inconvenience of delays and cancellations in inclement weather, those events do not pose a significant health or safety threat. There is however a very serious issue of rail's competing transport mode, (i.e. road transport) regularly impacting on the health and lives of those who don't even necessarily use the mode. We could take the howling of complaints when the sun comes out a bit more seriously if the more important issues were given more attention.