Orwn, the problem is well-defined, but my formulation of it, as it turned out, required clarifications (at least for this audience) - and that's fine.
But you are right that some other elements of the poem could cause associations with mirrors - and they wouldn't be helpful as clues for finding the right answer to my puzzle.

By the way, there is another aspect of the poem (forget now about my puzzle) which I think was essential for Usov (and which Gasparov, judging by his essay, didn't notice) - it is related to the phrase you paid attention to.

Namely, in the first stanza, Usov describes the environment of the translator: it is "motionless evening" (недвижный вечер), and he is talking (not exactly as in our translation) about the "disposition" (расположенье) of the original quatrain "in front of [him]" (передо мною).

In the "original" quatrain itself, it is also a motionless evening, but the words aren't used: it is "dusk" (сумерки) and "before night" (перед ночью) the "city is getting quiet" (затихает).

However, in the mock-translation we have: "evening hours" (вечерний час), "motion" (движение) in the city is getting quiet, as well as "in front of me" (передо мной), аnd "representation" or "image" or "reflection" (отображенье - а word quite consonant to расположенье from the first stanza).

So, I think the point Usov is making here is that the translation depends just as much on the original text as it does on the translator's own circumstances (mood, environment, etc.) We tried to reflect this in our translation.

Matt, I see that you and I have been thinking along the same lines, though you’ve gotten a little farther than I have with your speculative placing of the seventh reflection in the center. All I can add is that the middle reflection produced by two mirrors may be on the join or off it depending on the camera angle. Alexander told us so, but I didn’t understand until I looked at various photos. Now I’m going to repeat our question for Alexander.

Alexander, as intriguing as the poem and its translation(s) are, I feel they were a red herring in this puzzle (encouraging us to think of it as a word problem and based on multiples of two rather than three), but never mind that. I’m more interested in the solution. If I had a corner mirror to consult, this would all be crystal clear to me, but I don’t and it isn’t, so please treat me like a third-grader. Let’s talk about how many reflections there are (asking how many iPhones was another red herring). After some mental effort, I can visualize six reflections: two perpendicular mirrors produce three, and a third reflects those three (3 + 3 = 6). Or (and this is already twisting my brain out of shape) three mirrors can be paired in three ways, each producing three reflections, from which we subtract three shared images (3 + 3 + 3 – 3 = 6). So where does the seventh come from? Is it, as Matt suggests, in the center, with six reflections around it? If it’s simply an “up-side-down image below the actual one,” why does it appear only in the floor (or ceiling) mirror?

Carl, here https://math.berkeley.edu/~giventh/mirrors.pdf I posted three pictures.

The top one shows that if the angle between two mirrors is random, then the number of mirror images (produced by consecutive reflections in the counterclockwise direction) can be large, potentially infinite in the case when the angle is not a fractional part of 360 degrees but an irrational part of it.

The middle picture shows, in the case of two perpendicular mirrors in 2D, how the light trajectories from the object to the eye can be reconstructed from the straight rays from the three phantom images (color-coded).

The bottom picture adapts the photo that Roger sent to us to the case of three pairwise perpendicular mirrors, and shows one of possible (depending on the position of the observer) positions of the 8 images of
the candle.

As about the red herring: why?
I said from the start that it is a math, not word challenge.
The typical error one makes is thinking that 2 (or 3) mirrors produce 2 (resp. 3) mirror images - one per mirror.
Any poem consists of words, quatrains, etc., but this one is very unusual if not outright unique: it translates a text into the same language (and likens a translation to a mirror image). Moreover, the poem is translated into English, so that the initial text is reflected in two metaphorical mirrors.
Thus, my hint was fair: these two "mirrors" produce not 2 copies of the original text but 3. And the hint worked for Sarah-Jane!
As a surplus, there comes the fact that the poem refers to an actual mirror (that's how the metaphor is established), and so if you count the overall number of the Moons there, it is 8 - and for exactly the same reason why there are 8 images of an object reflected in three perpendicular mirrors. Namely, the corner of a room is 1/8-th of the whole space (accessible to Alice) and with every mirror reflection, it doubles - just as the number of copies (of the Moon, or quatrain) doubles with every reflection, be it physical or metaphorical.

Cool. The bottom picture is just what I needed. If it shows one real candle and seven reflections, then I finally understand. The penny has dropped, as the British say.

Ok, all I can do is speak for myself. If you had left the poem out and asked, not how many iPhones there are, but how many reflections, I wouldn’t have known the answer, but I would have known where to look for it. I ended up on a wild goose chase, but the chase was fun, and I learned something that I should have learned in the third grade, so all I can do is say thank you.

Carl

Welcome - and sorry that the post didn't quite work as an application of poetry to science. But it was all triggered by the title "Mirror Mirror Mirror" of Ralph's poem (so the fault is his :-)