Dear friends,
I find this discussion very positive and I was curious to read your point of view.
As we know, the eyes are easy to be "seduced" and coaxed, especially when it comes to vertical & horizontal lines or reparative patterns.
Long time ago I attempt an experiment and till nowadays I continue using this experience – a funny trick, when such music model appears – as the one shown by our gentle friend and "householder" OCTO:
"(Page 2) Here is another tricky question:
How to notate simultaneously these two measures? What rule will be applied, and what is the reason behind it?"
_________________________________________
I would explain my test, because as I mentioned, it works (at least for me).
Here is the sample:

- 01.jpg (126.34 KiB) Viewed 8857 times
On first glance, mostly on 5 mm
printed score, it looks convincing.
No doubt, we all know that the positions are altered.
On first staff (with G-clef) the intervals generate a perception as if the font is much heavier. (Please, find both sample provided by TISIMST and John – PAGE 2).
All the time when I see 3-ds I feel that the boldness of the note heads is an issue. Therefore in my custom fonts I always keep a narrow quarter note glyph for similar reasons (and for those, where the horizontal space is limited.)
The most important is to manipulate slightly and with care the distance between stems in order to look as even as possible.
Well, the note heads play also their role, but the eye follows the lines first.
In this sample the note heads on the first staff are between 90-95% narrower as a width compared to slot Nº207.
Second – the position of all note heads are moved individually to the left or right depending on stem direction and in relation to the usual stem's offset. As said, printed on paper, this effect looks almost invisible, but on the monitor is more prominent.
The third step: left-right slight "note movement" in order to equalise the stems in conjunctions with the bottom staff. The alternative position is "+" or "–" ¼ (even ⅓) of the normal stem and note position.
Here, on the next screen shot, you can analyse it using the grid lines.

- 02.jpg (193.95 KiB) Viewed 8857 times
It takes considerable time ended and that is the reason I perform this very occasionally, mostly when the distance between staves is bigger unlikely this one. I know, this is not perfect, but could help as an idea for test.