Isn't having a tag such as italian inviting questions requiring at least some knowledge of Italian? Isn't 'loan-words' sufficient? Aren't cross-language questions more appropriate on Linguistics?
An argument can be made (and I believe that at least one prominent EL&U contributor has made it many times) that the tags used at EL&U are completely useless and a waste of time. If so, focusing on which particular foreign language tag(s) should be dumped is rather like trying to to decide which particular deck chair(s) to throw overboard now that a collision with an iceberg seems inevitable.
My view of tags is that they can be helpful in noting particular areas of relevance within a question. For example, a memorable question posted in 2015 asks where, when, and under what circumstances the U.S. English term "pizza pie" originated. Since the word pizza by itself is of Italian origin, it would seem reasonable for the poster to have included a tag for "italian" among the other tags (which currently consist of "etymology," "american-english," and "food").
Although I am broadly interested in the English language's amoeba-like ingestion of words from other languages, I am particularly interested in instances where the appropriated words have undergone some degree of transformation in the course of their incorporation into the body of popular English. In such cases, having a tag that identifies the source of the appropriation as "italian" (or "yiddish" or whatever it is) may be quite useful to readers whose etymological interest doesn't extend to all source languages.
On a more general note, I think it is worthwhile to appraise the value of tags not in isolation but in the context of other tags that may appear in the same question. The fact that posters are permitted to include as many as five tags with a question suggests to me that the people who adopted that policy imagined that posters might use some tags to refine others—as might happen when someone uses a tag such as "italian" to provide greater specificity to a more broadly applicable tag such as "etymology."
My understanding, and use, of tags differs from what is suggested by your question. I apply tags after composing the question, primarily as a means of facilitating the attention of others who might be interested in answering the question, or at least interested in reading it.
Your question seems to be primarily an expression of your disinterest in etymology questions, or questions involving other specific languages but not centered on that other language; many questions tagged etymology might want a specific-language-other-than-vanilla-english (hereafter called slotve)content or subject tag, because that other language (or dialect; Scottish, for example, is sometimes considered a dialect) might be, or be said to be, a source, supply an etymon, etc.
So no, I don't think questions with slotve are "invited" by the tag; no, I don't think loan-words is entirely sufficient for everybody's purposes here; and no, cross-language questions about English etymology (among other subjects bearing on English usage) may not be more appropriate on Linguistics.
I seem to recall (through a veil of mild indifference) that it has been established on ELU Meta that there is some disagreement among ELU participants about whether or not etymology (or phrase-origin, etc.) questions are on-topic. This disagreement is sometimes reflected in overly aggressive close-voting by participants with little or no interest in, or expertise on, the topic. I also seem to recall that the perfectly good suggestion for dealing with these circumstances is for those with little or no interest in, or expertise on, a topic to recuse themselves from close-voting questions on such topics.
Whether or not I personally use a tag, have an interest in, or expertise on, a topic, I don't see attempts to do an end-run around disagreement about whether the topic is on-topic at ELU by foreclosing on tags that might be useful to those who do have such an interest, or systematically close-voting questions about the topic, as at all legitimate. Neither are such attempts, at least with respect to foreclosed tags, likely to be successful, because anybody with a certain reputation is likely to simply add the slotve tag back in if they think it's salient.
By way of example, I note that 10 of the 32 questions now tagged italian are also tagged etymology; 13 of the 80 questions tagged german; 9 of the 64 questions tagged spanish; etc. The translation tag also frequently co-occurs with slotve.