Well, is it?
A point on the globe can have multiple names changing through history (eg. Early Canadian provinces, or Holy Roman Empire to current Germany) but also sometimes at the same time (Derry City, and Londonderry City). They may have constant names at one level in the hierarchy, but other levels may vary - often with highly political overtones. In which "state" have we found Dublin City and how has that state been defined/named through say 1912 to 1949 (and by whom)?
On the face of the property box or diagram or report what do you want to show? You may want considerable variation to reflect either official usage at the time the record of the event was made ("strict transcription") or the perspective of the "reporter" (letter writer for instance) at the time.
But for geo-coding you need a name that is standardised to a form that the specific geo-coder can understand (which may be different to modern usage - e.g. previous posts about NYC, NY, USA etc.)
Pity the "standard" does not allow "Standard Modern Place Name" and "Standard Geo-code Place Name" - or is there a descriptor field somewhere in the GEDCOM spec that allows that?
) Although it's not the case here, I could well have found that I had both German and English language sources for that death and, short of recording multiple deaths for him, I have to make a decision between writing the English ("Prussia, Germany") or German ("Preussen, Deutschland") - and in my theoretical case, both forms are in the sources for the same event. I can't (in my mind) be precious about this, I have to make a choice, which will not, in this theoretical case, match one of the sources. This example says to me that there will always be instances where text from source and recorded event data don't match.