I have a degree in Chemistry and a Master's degree in Physics. I found the treatment of science, particularly physics, excellent. I think it could be difficult at times for a reader without a background of A level or equivalent. Never mind, it is possible to bleep over those bits and still enjoy the book. It hinges on different ways of seeing things and the consequent effect upon interpretation. Discworld, as in the three previous books in this series, provides the opportunity that valid laws of physics hold there. Full marks Sir Terry for achieving that.
As to the underlying story from Discworld: amusing but not comparable with the novels. I was rather disappointed that the relationship between the Unseen University librarian and Marjorie Daw, the Roundworld (Earth) librarian didn't develop further than a gift of a banana.
The conflict between science and religion is less satisfactory. I am a practicing Roman Catholic and personally I find no conflict. A point which is not made fully is that science is a rational system based on doubt and religion is irrational based on faith. These positions are not mutually exclusive; one can hold either, both or neither. The arguments against religion tend to be naive simply looking at God interfering in the physical world as required. I suspect that none of the three authors has religious faith (indeed it is specifically stated for one of them) and that is why they seem to have difficulty in really appreciating it. At times it smacks of "This is a hangover from the past where people were not as clever as us" I dumped that idea when I read James Cook's account of a meeting a Polynesian navigator in his exploration of the Pacific and realised that the man was as great a navigator as he was; even better considering he didn't have the sextant and Harrison's chronometer.
In short, there is no way to guarantee scientific truth; quantum mechanics and chaos theory will ultimately prevent that. Neither is there any way of proving religious faith. If you need another example; try proving, SCIENTIFICALLY, that one and one make two.