That would mean writing an error message whenever a non-ASCII character is used; this would be unhelpful for two reasons: a) in most cases, it would warn about a problem that doesn't exist, and b) since it would be shown so often, and because it would usually end up being wrong about there being a problem, it would tend to be ignored.
The reason for (a) is that it depends on how the machines involved are configured; if you type a Ü, it is encoded to an integer, and then later, decoded from that integer, back to a Ü, if the regional settings on the machine are the same as those that were in effect when it was encoded (I am simplifying things a bit here). If the regional settings are different, though, it will be decoded to some other character. In other words, it is not possible to know, when writing a file, whether it will end up being problematic later on, when it is read, because that depends on the machine being used to do the reading, not the one doing the writing.
We are working on things, here, to prevent this being a problem in the future (namely, changing all of Maxwell to use UTF-8 internally), but that work takes awhile. Even when we are finished, though, it will remain likely that there will be various softwares that you use, which have this limitation. There are only two ways of dealing with this: 1) making sure every machine that sees your files uses exactly the same regional settings, or 2) avoiding the use of non-ASCII characters. Ignoring the "moral" question of whether this should be something you have to think about, the fact is that from a business perspective, restricting yourself to using ASCII is likely to more reliably make you money, by reducing uncertainties in your workflow.
It is a bit like the decision whether to use something like Microsoft Word for your business documents, or to use plain text files; the former will be prettier, but will also likely cause problems at some point, when you run into someone who doesn't have Word, or who has an older version.
For more information on what is involved here, technically, you may like to read a bit about
ASCII,
Extended ASCII,
Unicode, and
UTF-8.