Principally I find the sentiment that there is a good question underlying this understandable, but I doubt we have the right to alter the question's scope unilaterally.
Part of the problem here may be rooted somewhat in the research standard, since in conjunction with the P.O.B. closure reason, it presents us with a catch 22 situation. If you require somebody to check references to prove that there is an ambiguity and preclude one from being used as the sole basis of an answer, then it is very easy to assume that once they actually do so that the answer can not be reference based at all, hence making it closer to a Primarily Opinion Based question. Unless we want to close all of the questions there has to be some leeway on one side or the other, and I am not so sure that English Language & Usage's treatment of the network wide subjectivity standards and the relevant closure reason entirely accurate.
As a result of the War of the Closes, the Not Constructive closure reason was changed to Primarily Opinion Based, and eliminated language regarding polling, debates and extended discussion. The change suggests that we need not have a certain answer, so much as an answer that can be coraborated. There are two proofs of that. The first proof is that there is good subjective in Good Subjective; Bad Subjective, and the second is that the language which remains is all based on types of corraboration that can be proferred:
Many good questions generate some degree of opinion based on expert experience, but answers to this question will tend to be almost entirely based on opinions, rather than facts, references, or specific expertise.
Moreover, it should be noted that references are not the only type of corroboration in that. We could try to argue a case based on facts, or relevant expertise if no adequate reference exists. Relevant expertise could come from a border patrol agent, a newspaper reporter, a legislator, a prisoner of war or just about anybody who is likely to have experience with the word and willing to disclose any relevant qualification which enable them to speak on the particular subject with greater authoritative guidance than the average person.
The other, and perhaps much more convincing way way to do so is to examine the facts of usage. It is in our very name folks. We could be examining collocates. A cursory example of that is looking up good concentration camp, evil concentration camp with the Google ngrams tool.
Surprisingly, no singular concentration camp is described as evil according to the chart, but even more tellingly is what you get when you check the book results from the 1970s through 1991. The first page of results show that the word good is either used ironically, or with its particular sense of effectiveness, rather than benevolence.
From page 5 of the 1973 United States Congress Senate Judiciary Committee we see the report U.S.S.R. labor camps: hearings before the Subcommittee to Investigate the Administration of the Internal Security Act and Other Internal Security Laws:
When his father was arrested, the K.G.B. tried to press his mother to become a secret agent by telling her that, if she provided good information father was arrested, the KGB tried to press his mother to become a secret agent by telling her that, if she provided information about neighbors, his father would be sent to a "good" concentration camp. His mother refused. His father was sent to Kolyma Complex where he died after ten years. He was rehabilitated posthumously in 1985.
Putting what a "good concentration camp" is in perspective is this little snippet from The Grand Tour by Jerry Herman, Michael Stewart and Mark Bramble - 1980
Either they put me in a good concentration camp or a bad concentration camp. If it's a good concentration camp that's not exactly good but it's bearable … but if it's a bad concentration camp there are still two possibilities …
Black education and the inner city by Marva Collins et all (the names are blurry and hard to make out) for the Lincoln Institute for Research and Education
Lincoln Institute for Research and Education (1981).
If there is a lot of noise going on in room 5 where Miss Wilcox is in charge, she's likely to be in trouble. She will be told that she's not able to maintain discipline. The idea behind this ancient motto is that a good classroom is a quiet classroom. I would like to revise that and say that a quiet classroom is a good "concentration camp."
If we were strictly being asked to arbitrate between the Oxford and Merriam-Webster definitions, then I would suggest that this evidence favors Oxford's, although perhaps a more comprehensive look into the matter is merited than the cursory check I provided as a proof of concept. This particular collocate is also perhaps biased towards this particular finding. Why would we describe internment camps in terms of good or evil, for instance, unless we considered them a moral issue?
Another point in the question's favor, in this regard, is that it specifically requests objective evidence of some sort, so at the very least, not all answers are "equally valid". Some will be better than others dependent on how convincing they are. This is not a What's Your Favorite …? Type question.
However, the question being asked is not our own and its scope is not ours to reshape. The ultimate principle behind editing a post is that ultimately, the poster has final sayso over the contents of that post. Stack Exchange has a greater emphasis on individual post ownership than Wikipedia, which is part what makes it a more appealing format for making contributions than Wikipedia. See The Great Edit Wars for reference. Editing guidelines specify that we should edit to clarify the meaning of a post without changing it.
We can suggest improvements to the question through commentary, but ultimately the goal of a question and answers website is to answer people's legitimate problems, and so ultimate arbitration regarding what constitutes the best answer to the questioner is left to the questioner, with answer acceptance, which will bring any answer the questioner likes best immediately to the top, even if that is judged on an entirely nonlingual basis.
Right now, I think the core issue is distinctly outside of scope. We may be able to tell you which meaning of the word is better, or what terms are more accurate to describe a particular concept, but if Oxford is correct, how do we evaluate whether the Mexican Internment camps are adequate facilities for their purposes using standards of evidence relevant to our field? Whether or not a person is being treated humanely is a subject that we can not determine using words or syntactical structures alone, and the question in its present form solicits that information in particular. we'll have more of a political debate than a lingual one. That not only requires us to know the conditions of the camps, but evaluate them in terms of how adequate those facilities are equipped to do their job, perhaps relative to other facilities in other nations. Why should a grammarian be expected to know any of that? Any answer to the question without that information will be incomplete, and anybody answering this question will be tempted to address that issue, leaving whatever legitimate linguistic basis this question might have as a secondary factor for determinations made by luchonacho.
If the questioner would like to change the title of the question for us, with an understanding that we will not make that particular determination, then sure, we should reopen it. If not it should probably stay closed. Otherwise I fear that the question may constitute a political "rant in disguise" that lacks "a constructive, fair and impartial tone towards determining the subject-matter of interest", against our subjectivity guidelines and that the final determination will be based moreso on personal opinion than fact.
If the original poster edits the title to something along the lines of Are Inhumane Conditions a Necessary Part of What the Term Internment Camp Means? then I would be glad to cast the question a reopening vote, but otherwise I think I shall decline, and advise that somebody else make a more focused version of the question instead, with reference to the original and its problematic framing. We are unable to really be sure what a question aims to achieve beyond the questioner's own input.