I don't see how cosmopolitanism implies wanting a single, powerful governmental entity to literally control the world, though. For example one can very easily be an egalitarian with regards to people of various cities, or ethnicities without thinking that there should be one global city with a world mayor or a single homogenized ethnicity in the world.

Justifications for nationalism could be pragmatic (e.g. out of other possible systems... let's say hypothetically, a world of city-states or millions of tiny tribes, or a world of a single hegemonic empire etc. the world we have now with nations of the sort we have now works decently, and perhaps trying to reorganize things too much at least in the short term might be bad for society) and that's a fair justification even if it can be called out as "status quo is good" bias. But we have to be careful to not conflate pragmatism or politically, logistically and legally convenient ways of arranging people with moral or value judgements of how to arrange them in terms of our circle of concerns. 

Also, it might have not been your main point of the metaphor, I guess another difference between the corporation example is that in a world of decentralized corporations and decentralized nation-states alike, people have nonetheless a lot more say in moving between the former than the latter and people's fates are much more strongly tied to the latter as they stand now in an unchosen way. This has strong implications for how to help out. 

If someone is doing poorly because their job or livelihood's tied to a poorly performing corporation (for e.g. if they live in a company town or place dominated by one industry), we'd advise them to quit or move, the altruistic thing is to get them to join a better one, perhaps letting the old one "crash and burn". We'd not often assume they'd stay tied to it while we channel money to those poor workers, keeping them and their corporation afloat. On the other hand telling someone "if your nation is faring badly, just pack up and move to a better one and let the old nation fail like a company going bankrupt", is not feasible advice (and many people would not want it). Thus, because we're "stuck" with people in nations who cannot easily leave/change their situation, thinking about how to help people in other nations who cannot easily leave problems behind becomes important.  

For nations where it's often accidents of history (or family) that determine citizenship, stuff like quality of life, average wealth differing by orders of magnitude, safety, health, access to a good life overall etc.,  - yes people can emigrate/change nations but it's costlier than changing jobs, companies, neighborhoods etc. - it becomes more like an unchosen trait that some people were just lucky or unlucky to have and unable to do things about.  I guess this kind of relates to how people with more "universalist" mindsets often feel that discriminating based on unchosen traits -- e.g.  ancestry, race, sex, and other things marked down from birth, who one's parents are etc., is wrong. Impartiality suggests that we not treat these as morally significant, It's not always a rule of thumb -- e.g. after all many things like personality, intelligence, etc. that make us treat others differently are unchosen (and in some philosophical way, I didn't "choose" to be a human and not say a worm), but it is an ethos.

I guess the question is, is internationalism something worth pointing out or having a concept for, or is it just a byproduct that naturally follows from"impartiality"?

After all,  many boundaries are crossed as a consequence of some degree of "impartiality" (e.g. the act of people helping others might cross family, social circle, neighborhood, town, city, region, province/state/territory/intranational division, class, ethnicity/race/ancestry, hemisphere, time zone,  generational, and in the case of animal activists, even species, boundaries). 

But it's not like every time any of these boundaries  (which in principle are too numerous and too arbitrary to count, even if some are enforced, legally, socially etc. and others not so much) are crossed by would be do-gooders, they get flagged as noteworthy (e.g. we don't always call out helping people in a different city, inter-regionally or inter-ethnically as too strange, unless, perhaps you're living in a society of strong cross-city, region or ethnic tensions and then, crossing over to help the "other" marks you as unusually heroic or a "good Samaritan"). 

Naming "internationalism" as a special kind of boundary-crossing would seem to imply it's a particularly significant one, among many others, because the barriers are exceptionally high. 

It seems to me that "internationalism" would be seen as "a thing" precisely because nationalism is so strongly the default or status quo and nations often an unquestioned unit of organization (you could flip the frame of "internationalism is only a tool" to "nationalism is a tool" for doing things, good or otherwise, after all).