Sure, there's some bound to the list of characters reasonable to choose, but Ricci's choice has cemented certain old-fashioned choices which are no longer updat

Sure, there's some bound to the list of characters reasonable to choose, but Ricci's choice has cemented certain old-fashioned choices which are no longer updated (see e.g. 细). 欧罗巴 indeed is from Europa, but it has persisted as the full name of Europe in Chinese (欧罗巴洲) as opposed to its usual abbreviated form (欧洲). Indeed there is a series of official handbooks published by Xinhua for a variety of languages dictating how the phonemes in those languages should be transliterated into Chinese. Ricci's translations do not follow these. Having approximately 400 years of priority lets you do that. The handbooks also generally defer to well-established, preexisting names and conventions where they exist. Also to answer one of your questions, no not all characters in Chinese have meaning. The more well-known case is purely phonetic characters, although it is quite tricky to find ones that have absolutely no other attested meaning. One example of this is 噶. The less well-known case (I don't know whether the phenomenon even has a standard English term) is of indivisible or "continuous" polysyllabic words, although these words are not always obscure. These are the exceptions to the rule that a Chinese character is a morpheme. Here the constituent characters of the word have no meaning individually and exist only in the context of this word. They are not reusable linguistic units outside of that word. Some, but not all, of these later evolved to have independent meaning to fit in with other Chinese characters.