I found this a useful and illuminating book, albeit one that at times was dry or with claims that were less than perfectly supported.
I suspect most scientists would agree with Montgomery's assertion (finally given on page 175) that yes, science does need a global language. It's prima facie obvious that this is helpful for a field that is so incredibly dependent on mutual intelligibility and that now involves people from all over the world. Montgomery shows how this global language has been helpful for modern science and is organically spreading on its own due to people recognizing its utility (independent of present or past imperialism by the UK or US).
Indeed, science has almost always had an international language connecting scholars. Montgomery goes through several examples in some detail, including Greek, Latin, Arabic, and Chinese. He shows how important these languages were for scholarship, even when they were dissociated from any vernacular usage. For an example from my own field, a millennium after Latin was no longer used by most people in Europe, it still allowed Kepler (German) to work with Tycho (Danish), read Copernicus (Polish), write to Galileo (Italian), and later be read by Descartes (French) and Newton (English).
Montgomery thus shows that the period from around 1700 to 1980, where there was no one language of science, was a historical aberration. And indeed it was possible only because science was fairly concentrated (first only in Western Europe, later extended to the US, Eastern Europe, and Japan). Even during this period there were only a few languages of science—mostly not more than three—and little enough publication that results could be translated.
Today, we have people contributing to science from all over the world, with hundreds of native languages (at least), and more scientific papers published every year than were published in the whole of the 19th century. This demands a single language.
English has become a global language in other realms as well, including in business, diplomacy, entertainment, but its widespread usage may be most prevalent in science. Montgomery emphasizes the point first made in the 1990s that English is the first global language. Empires have spread languages across large regions before, but never the entire world. So its future is unpredictable, but if the past is any guide, it will remain the language of science for centuries (even if the US itself crumbles).
I have a few quibbles with this book. First, Montgomery uses bulleted lists too much, sometimes including things on them he says he will address later but doesn't. It feels like we got a mostly finished draft of a book, with some parts still in outline, rather than a truly complete book. Second, he gives several example passages of "nonstandard" "world Englishes", most of which don't actually seem to be truly nonstandard, and could just be style choices. Certainly Montgomery's own use of commas and asyndeta doesn't always conform to my own standards, despite us both being PhD-educated Americans. (To be fair, some examples he cites do vary from standard written English. These do help make his point that English no longer fully belongs to the anglophone nations, and some future developments will take place outside their control.)
Third, the editing of this book wasn't that great. Minor mistakes with major implications, which should have been easy to catch: at one point he says "Eastern Roman Empire" when he means "Eastern Islamic Empire", at another point he says "1.5 people" when he means "1.5 billion people", etc.
A fourth issue I have is with Montgomery's prescriptivist claim that "intelligibility must be maintained, but this does not argue for rigid standards" (p. 178). I take his point that natural language development and inclusion of nonnative speakers should make us more accepting of some nonstandard usages. No papers should be rejected for linguistic awkwardness, particularly given that we're trying to include contributions from colleagues from countries without well-developed English teaching. Indeed. And yet, he does not acknowledge that intelligibility is a continuum, not a binary. Someone fluent in English in 2022 could read three passages from about the same time: let's say something from Joyce's Ulysses, a poem of E. E. Cummings, and a New York Times article on the election of Calvin Coolidge. It's undeniable that it's far more difficult to make sense of the first two than the last. If two scientific papers are the same length and topical complexity, the one that takes four times as long to make sense of has a problem, particularly if some parts have ambiguous meaning. In fact, this is particularly important when some readers aren't native speakers. Lapses from standard English by a Korean may be less intelligible to a reader from Mozambique expecting a certain standard than a reader from the UK whose experience encompasses a fair variation in what is understood as English. (I've experienced this with spoken language, when I was able to understand a particular British accent that was entirely impenetrable to my French friend, despite his masterful fluency with English.) Completely rigid standards of writing are not helpful (or even possible), but mostly rigid standards are.
My last issue is that Montgomery argues poorly when he advocates against anglophone monolinguism. Earlier in the book he suggests we'd have to resolve the Sapir–Whorf hypothesis if we were going to argue that multilingualism is really helpful, but then he never mentions it again. He spends essentially the entire book arguing that skill in English is now becoming a basic skill, akin to math, and should not be viewed as a normal foreign language issue, but then does a quick and essentially off-topic dance at the end to argue that foreign language acquisition is good for native English speakers. The primary reason he gives is cultural competency, which is a bit bizarre. I suspect he's arguing for language acquisition in an attempt to insulate himself against attacks that he's advocating for his fellow anglophones to be lazy. (He's saying everyone else should learn English, so for the sake of fairness we anglophones must all learn another language. Understandable, but not actually a sensible argument.)
Now, I personally believe that everyone should be multilingual for the intellectual stimulation and mental broadening, and it's actually a bit of a disappointment to me that it's been hard for me to acquire and maintain other languages due to being so firmly ensconced in a monolingual region. Unlike most places in the world, where I live in the US has no non-English-speaking communities for over 1000 miles around it that are not insular, marginal, and not set up for outsider visitation (e.g., an immigrant community in a large city). Friends of mine with other first languages are all also fluent in English, so our choices are broken communication in their language or easy communication in mine.
That said, I've found it enriching to study a few languages, and one well enough to get by. However, it has never been very helpful for me to know French for doing my science. I've been to multiple conferences in France, and while it was nice to know French, I never actually needed it. All communication with French colleagues was done in English, because their English was better than my French (and collaborators and co-attendees at conferences were Korean, Italian, German, and so on—English was the only common tongue). I've never read a single paper in French; the prominent astronomy journal published in France (Astronomy & Astrophysics) is solely in English. I've read a single technical book in French (a third edition, when only the first edition was translated to English), and while it was well written, numerous English-language books covered all the same topics very well. I got job offers in the Netherlands and France without ever communicating in anything other than English with my potential employers.
To be fair to Montgomery, as a geologist he's done a lot of field work. For scientists in the field, it can be useful to pick up some other languages, particularly when doing intense work in one particular region, or when working with less educated populations. Jared Diamond would not have gotten far in field work in Papua New Guinea without picking up some local languages.
But making a nonspecific language a requirement of school or grad school, as Montgomery suggests, is less well motivated. I could imagine an argument for it, but it's not one related to doing science, and thus not relevant to or motivated by anything in this book.