I have undertaken repeated extensive rewrites of the question, each time in response to actual moderator criticism. At this point, the answer as it stands is quite confidently within the bounds of the question itself, irrespective of the interpretation of "what the question is actually asking", due to the specific wording of the question.
In particular, the question asks, within its body,
A quick google search returns several articles such as this one [referenced] claiming that airlines make less than $10 in profit from each customer
The wording, "profit from each customer", carries a different connotation to "profit per customer" as found in the question title.
This distinction, this connotation, is the root of the questioner's concern about whether the claimed number is correct. By only answering with the literal "total profit divided by total customers" evaluation, the question is being answered incompletely, thereby giving a false understanding of the situation.
Also notice that the questioner had asked, at the end of the question body,
Do airlines take losses on some flights and make up for it by turning a profit on other flights? Is there some other rising cost to account for?
These questions have since been removed by a moderator, explicitly in an attempt to make my answer invalid, as they are a fundamental part of the question being asked. The moderator's reason for the edit was literally "Removed speculation leading to answers that don't address the claim." - despite the fact that it does address the claim.
These questions further demonstrate that the raw claim of "airlines make $10 profit per customer" cause misunderstandings that are at the root of the question being asked.
This site stands opposed to misunderstandings such as this, and by allowing answers to stand that address only the literal "is this true" and not "is this fact being understood correctly", you end up only contributing to the problem.
To see what I mean, consider this answer I gave to another question. The literal question is "does this actually appear in the Talmud?", and if the answer were purely literal, it would be "Everything except that last sentence is in the Talmud, the last sentence is a fabrication"...
But it's not just about whether it's technically true - it's also about whether it's accurate in terms of context. If someone claims that the sky is blue, normally it would be fine. But if they said it during a sunset, they'd be wrong - the context matters.
And so, in that answer, which is both accepted and significantly upvoted, I pointed out that (aside from the fabricated ending) it's a literal interpretation where the literal interpretation is misleading. Without that part, the answer would also be misleading.
Indeed, it's standard practice to answer not just "is the literal claim true", but also "is the claim valid" (in the sense of whether, within context, it's a misleading claim couched in a factual statement). This can be seen, for example, here, where the moderator who originally took issue with my answer posted both a question and an answer. The notable claim was
The findings, to be published in the November edition of ‘Personality and Social Psychology Bulletin’, concluded that men find the propect [sic] of dating intelligent women intimidating.
This, as demonstrated in Sklivvz's answer, is a true statement. And it is - a study did indeed find that men find dating intelligent women to be intimidating.
By the reasoning applied by the moderators to my answer, this should be the end - a study found it.
However, Sklivvz's answer then went on to explain, without any references whatsoever (the only reference in the answer was to the paper), why the study itself shouldn't be considered reliable, with assertions such as that removing students who had figured out what was being tested was a "systematic bias".
More generally, it's not at all uncommon for "it depends on how you interpret the question" to be an answer. For example, Oddthinking's second-highest-voted answer is addressing the question of whether children were safer in the "good old days". His answer involved five different interpretations of the question, each of which produces a different result.
This isn't unusual - many notable assertions make use of plausible interpretation to be technically correct while also being misleading. Oddthinking's answer could have provided only interpretation 5, which would imply that children were noticeably less likely to be victims of child abuse in the past, and thus the claim is true by technicality.
But this is Skeptics. The whole idea is to cast doubt not just on the literal claim, but on its legitimacy as a claim.
In the case of my answer, I am pointing out that, although it's technically true that airlines make $10 per customer in profit, this is a misleading claim that causes confusion such as that shown by the questioner. While the literal claim had been addressed in another answer, I judged the literal claim to be misleading, and thus provided an answer that ensures that the claim isn't misunderstood the way that it was misunderstood by the questioner.
It appears that moderators are actively and spitefully trying to delegitimise my answer, going so far as to change the words of the questioner with the sole aim of justifying the deletion of my answer. Perhaps I'm mistaken, and they honestly believe my answer doesn't contribute - in which case, it would be helpful if they provided justification for why my answer isn't relevant (and not just asserting that they disagree with my arguments, without evidence), rather than just using their mod powers to actively try to stop me from providing it.
Given the above, I ask that my answer be undeleted, as it is relevant, it is backed up with references, it is entirely within the theme and point of the Skeptics site, it contains no "original research", and it is on-topic even after the removal of parts of the question.
One more demonstration of my point. Consider this question, about whether it's possible that a monk could go 9 days without food, water, or sleep. There are two answers. One answer talks about what Buddhists mean when they talk of fasting, etc, and never once comments on whether the referenced claim:
A Buddhist monk on Wednesday finished a grueling nine-day ritual of not eating, drinking, or sleeping as he chanted sutras 100,000 times. [...] He was the 13th monk to complete the test since the end of WWII.
is correct. The other answer addresses the plausibility of the claim, by using references to show that any one of the three, alone, is highly unlikely - going 9 days without food, or without water, or without sleep, is unlikely in the first place. Then it notes that the combination would make things even less likely.
By literal interpretation of the question, the second answer is the right one, while the first answer is off-topic for talking about the context of the question instead of the question itself.
One of the answers has been flagged while the other has received more upvotes.
The flagged one was the second one, accused of being "theoretical", despite it being the "literally" right answer. This is the normal logic behind acceptance of answers - the answer that addresses the claim in context is more valuable than the one that answers it literally. By the same argument given by Oddthinking, "what does it mean when a Buddhist claims to fast for 9 days?" belongs on the Buddhism stackexchange, and thus is off-topic here.