One area where this is important is that of historical linguistics, because the terms are often confusing and there is no overall system for nomenclature, at least for older terms that are still central to the discourse (say like in organic chemistry). For example, the names of some branches of the Indo-European languages have strong geographical connotations. "Iranian" speakers live in Iran (roughly speaking, the historical Persia) and Indo-Aryan speakers reside in the Indian subcontinent. But these appellations cause problems when one speaks of the ancient Scythians, who ranged on the Black Sea Steppe, or their successors, the Sarmatians, who eventually settled upon Hungarian plains. Both are often classified as "Iranian" because of the features of their language. But the term elicits in the mind of the target audience outside of linguistics the geographical term, so many assume that the Scythians and Sarmatians swept out of Persia via the Central Asian steppe or up through the trans-Caucasian plains. An even more vexing issue is that of "Indo"-Aryans. Several decades ago an Indo-Aryan group was found to have settled in the kingdom of the Mitanni, roughly the upper regions of the Euphrates now within the borders of Syria. There were terms within the predominantly Hurrian language of the Mitanni, which was non-Indo-European (and not Semitic either, it was one of the many hard-to-classify languages which existed prior to the recent expansion of both the aforementioned language groups), which were clearly Indo-Aryan. For example the number terms that defined the laps a chariot made were derived from Indo-Aryan numbers. Additionally the gods sometimes referred to within the treaties were Indo-Aryan. The key point is that the linguistic clues suggested an Indo-Aryan association, not an Iranian one. When you explain this to a lay audience often the first response is that someone how a group of Aryans traversed Persia from their homeland in the upper Indus valley and settled in Syria. But there are problems with this hypothesis, because the linguistic fragments show no evidence of familiarity with terms that are distinctive to Indo-Aryan due to the encountering of objects and creatures local to India. To top it off, the Mitanni dialect exhibits archaisms that suggest it predates the Sanskrit variant of Indo-Aryan found in the Rig Veda. This is plausible since the Mitanni tablets date from 1600-1500 BCE, and at this point the Indo-Aryan dialect was likely used for ritual or formalistic purposes and so preserved a more ancient manner of speech.2 The Rig Veda was certainly fixed after 1500 BCE, though before 1000 BCE, and its language was a living tongue which was still evolving.
The "solution" to this mystery is rather simple, it seems likely that both the Iranian and Indian Aryans derived from what is termed the Andronovo Cultural Complex, which existed in the late Bronze Age around the Caspian steppe and further east into northern Central Asia. When the original Indo-Iranians dispersed from this region it is likely that they spread out in multiple directions, and there was already some differentiation between the "Indo"-Aryan and Iranian tribes prior to this dispersal.3 Some of the Indo-Aryan groups settled in India, and gave rise to the languages spoke by 3/4 of modern Indians. Others seem to have become absorbed into the milieu of the Middle Eastern cultures, disappearing from history. The Iranian speaking groups eventually dominated the Persia plateau as well as the Central Asian river valleys, but, some of them also migrated to the steppes to the north of the Black Sea and further west. Because linguistic distributions are a palimpsest these patterns and migrations have been obscured by the spread of Turkic languages in Central Asia (with Tajik and a few other Iranian languages as holdouts), breaking the continuity between the southern and northwestern Iranian tongues (Ossetian is a relict in the Caucasus of the western Iranian dialects). The extinction of all Indo-Aryan dialects outside of India also has resulted in the fact that that clade of the Indo-European languages is modified by the term Indo, when prior to the historical period its distribution was possibly far less geographically constrained.4
The same caution extends to many terms which have geographical origins, the classification of "Italic," Latin and its derivates + all the Indo-European non-Latin languages (Umbrian, Oscan, etc.). Or "Iberian" for the extinct language of the Tartessians of southern Spain, which might have a relationship with other dead languages of Western Europe or North Africa.
I have placed a small map for illustrative purposes below the fold.
1 - Philosophically this was a view espoused to some extent by the later Wittgenstein and championed today by many "Post-Modernists." I believe that modern cognitive science has falsified this view.
2 - The preservation of Mitanni Indo-Aryans terms relating to horsemanship is not surprising since it is hypothesized that Indo-Europeans introduced many elements of horse culture into the Middle East. As a point of comparison, Latin was preserved in Byzantine culture the longest in the military and the legal profession, two areas where Western Roman culture could compete with the Greeks.
3 - This idea of pre-dispersal differences and identities for various groups is a neat solution to why the Tocharians, the Indo-Europeans who settled along the northern rim of the Tarim basin in modern Turkestan (it seems likely that the southern rim of the basin had an Indo-Iranian population) are classed with the "western" centum clades of Indo-European, Celtic, Italic and Germanic, as opposed to the "eastern" satem groups, Greek, Armenian, Balto-Slavic and Indo-Iranian. Though some scholars dispute the salience of the centum-satem distinction, other points of evidence do suggest that there was an association of the pre-Tocharian tribes with groups that later founded the western branches of the Indo-European language family (in particular the Celtic branch). This association likely occurred in the Proto-Indo-European homeland, possibly the grasslands of southeastern Europe and north-central Asia.
4 - One model holds that in fact the Persian plateau was dominated by Indo-Aryans, and the Iranians were latecomers who divided the continuity of Indo-Aryan groups which settled in India, Persia and the Middle East. It is interesting to note that the archaic Indo-Iranian languages, Sanskrit and Avestan, tend to exhibit an inversion of some terms, for example Indo-Aryan daeva has positive divine associations, but in Iranian it is a negative term (hence, devil). The same inversion is found in the term asura, a race of anti-gods in Indian mythos, but on the side of the good God in Iranian tradition.