I guess what I'm wondering is how obscure it has to get (or perhaps how low the GDP of the people speaking a language has to be) before custom spell checking rules aren't considered worth it to bigcorps anymore. Though I'm also kinda interested in what this weird language thing is you're talking about.
Microsoft has an okay Finnish spellchecker. It's really a shame no Finnish university has developed an acceptable open source Finnish spellchecker though AFAIK.
Word-list based checkers actually work for English, which gives an impression that they would work for all types of languages. For agglutinative languages (which I think is about 50% of world languages) not so much, since the list of valid words is basically infinite.
But combined with a high-quality root word list it's possible to write a computer program that returns if a given input word is a valid form of some known root word, i.e. a spellchecker.
In English, you can think the problem same as writing a spellchecker that can tell if an input word is a valid chemical name for a molecule, like 1,3,7-trimethylxanthine (caffeine) or adenosine-5'-triphosphate (ATP). Clearly, we can not write an exhaustive list of valid words here. But it's pretty easy to write a computer program that can tell if the input word at least seems like a valid chemical name devoid of typos.
I guess what I'm wondering is how obscure it has to get (or perhaps how low the GDP of the people speaking a language has to be) before custom spell checking rules aren't considered worth it to bigcorps anymore. Though I'm also kinda interested in what this weird language thing is you're talking about.