In the US, “south asian” is fine. In the UK, there’s enough of each group where I think you should distinguish Indian versus Pakistani versus Bangladeshi. It’s not a group of people who want to be lumped together, same with Koreans, Chinese, and Japanese. They're also far less closely related to each other than say British people and French people.

> I think you should distinguish Indian versus Pakistani versus Bangladeshi. It’s not a group of people who want to be lumped together, same with Koreans, Chinese, and Japanese.

It is a very different situation though. 100 years ago, there was just one country, India – with a great deal of internal ethnic, linguistic, and religious diversity. And then, it was split into two over religion, and then one of those splits split again (due to a whole host of issues, although ethnicity and language were both significant factors). Whereas, China, Japan and Korea were distinct countries going back many centuries. You aren't comparing like and like here.

> They're also far less closely related to each other than say British people and French people.

In ethnic and linguistic terms, the majority populations of both Dhaka and Kolkata are basically the same people, the big difference is their religion (although Kolkata is 20% Muslim, and Dhaka around 3-4% Hindu). One can make similar points about the relationship between Lahore and Ludhiana. It is true that people in Dhaka and Lahore are not closely related, but the same is true between Kolkata and Ludhiana. The case of the British and French doesn't have any real parallels.

India was never one country until the modern incarnation. It was centrally administered under a series of empires, like the Austro-Hungarian empire. The groupings within the subcontinent are more different from each other than Western European countries. Even Bengalis, a distinctive ethnolinguistic group, are divided between India and Bangladesh based on religion, which is a major division—a much bigger one than the religious divisions that differentiate countries in Europe.

Many countries we identify as distinct in Western Europe are much more similar to each other than India, Pakistan, and Bangladesh are. For example, France and Spain—both are Catholic, and their languages were considered merely different dialects of Vulgar Latin until the 1400s.

> India was never one country until the modern incarnation. It was centrally administered under a series of empires, like the Austro-Hungarian empire.

At the height of the Mughal Empire, it encompassed most of what would later be considered India, with the exception of the far south. You can argue that an empire is not a country, but if we argue in that way, well for most of European history, Europe didn't have countries either.

> Even Bengalis, a distinctive ethnolinguistic group, are divided between India and Bangladesh based on religion, which is a major division—a much bigger one than the religious divisions that differentiate countries in Europe.

There is a significant minority of Bengali Muslims in India, and a smaller minority of Bengali Hindus in Bangladesh. The Indian-Pakistan and India-Bangladesh divides have never neatly followed the religious divide, certainly never as neatly as religious nationalists on both sides would have it.

> For example, France and Spain—both are Catholic, and their languages were considered merely different dialects of Vulgar Latin until the 1400s.

Old French and Old Spanish were already identifiably distinct languages by the 9th century. Some argue (following Robert Wright) that a major impetus for them becoming identified as distinct languages was Charlemagne's Latin pronunciation reform. Prior to Charlemagne, Latin in Romance Europe was pronounced based on the local dialect – so the same Latin text would be read aloud using a French pronunciation in France, a Spanish pronunciation in Spain, an Italian pronunciation in Italy, etc. Monks from Britain and Ireland, having no indigenous tradition of Latin pronunciation, instead devised their own based on phonetic spelling. Charlemagne's advisors convinced him (however incorrectly) that the British/Irish tradition of pronouncing Latin was truer to the original, and so he imposed it on the Church–which suddenly made Latin and the vernacular seem like much more different languages. Now, that said, Wright's thesis is controversial – but even its academic critics, I don't think many of them would agree with you in putting the distinction between vulgar Latin and Old Romance so late.

And, while no denying the difference between Islam and Hinduism is bigger than difference between Catholicism and Protestantism, Europe shows parallel cases of countries with a common language and culture, where the major difference is the dominant religion – the primary difference between Austria and Germany, is the former has traditionally had a Catholic supermajority with a small Protestant minority, the latter was traditionally dominated by Protestants, albeit with a large Catholic minority.

The vulgar of different European places was already splintered into very different variants many centuries before the 1400s, it just matured from a continuum of random informal dialects, the further you travel the better linguist you have to be, to more consolidated and more recognized languages (plenty of different ones in "France" and "Spain").

One reason this American (me) says "South Asian" is because here "Indian" also means "indigenous North American".

If hypothetically the government of India were to declare that the name in English of the country shall in the future be "Bharat" or "Hindustan", Americans would rapidly switch to the new name, and I would no longer be tempted to say "South Asian" when I mean Indian.