I don't have access to the full article, but it sounds like they didn't examine the sliding scale of height preferences, by one's own height.

The 21cm vs 8cm is the mean, and their sample size was large enough to be statistically valid.

I did specifically quote the part that includes 'best satisfied'.

Ie, the 'ideal' partner height.

Many people often choose partners that are ... close enough to many ideal traits, weighting them in different and complex ways, often not even entirely fully concsious of the nature of how they weight or order their preferences, but thats all way outside of the scope of this paper.

Yep, its possible the uh preference differential changes as you approach extreme ends of height, but the problem is that, being a statistically representative sample, it doesn't include many people who are very short, or very tall.

Anyway:

https://annas-archive.org/md5/50413a744e4887cff238a542b59b19b2

Here's the whole paper!

But from the publicly available text, it doesn't seem to report on whether that preferred delta between one's own height and the ideal partner height changed with the absolute height of themselves.

Yeah, that seems to be my take away as well, they don't go into precisely that in the paper.

Or is there a sliding scale where already tall people aren't exactly looking for excessively unusual outliers, and that the preference of tall women is something smaller than 21cm, such that the overall average might be that very short women prefer a big height difference but very tall women prefer a small height difference?

Apologies for shit tier resolution, I am on mobile:

'Female' meaning, the male-preferred height of a female partner, 'Male' meaning the female-prefered height of a male partner, so that... may be backwards depending on your inution for reading graphs.

Also these are 2 SD bounds, 95% CI, I kinda cropped out half the text that explains that, whoops.

So, yes, this effect you mention does exist... but they do not seem to focus on it in the paper.

Unfortunately, I am not seeing a visualization that or equation that more specifically and precisely answers your question of whether or not very tall or short men or women are less uh, height-choosy.

Perhaps I am missing it?

....

Here's another way they visualize their data:

Now, here, 'Men' means men, 'Women' means women, and the x axis is [male height - female height].

So, very broadly, yeah you see that the sort of mutual sweet spot of both partners being decently satisfied with the height difference is roughly a man being 13ish cm, roughly 5 inches taller than a woman.

So, from that, maybe 'Short King' has a realistic shot with 4'10" women, not 4' 7"?

???

...

We can also see that women's satisfaction with a male partner's height uh, nose dives as a women is asked about a man who is going from 13 cm taller than them, to the same height as them... but then does rebound once the heights are just inverted.

This is also the only situation where the man is less satisfied with the pairing than the woman, on average, (untill you get to men being about 18cm taller than the woman, then its roughly the same gap as the height difference increases) but the men have huge CI intervals in this instance, indicating many men actually don't mind this much, and some men mind it extremely.

Meanwhile, women generally dislike being taller than their man, with a yes, expanded CI range, but far less than that of the man, indicating that this is a less variable and more common ... anti-preference for most women, in general, than it is for men.

...

Somewhat oddly, to me at least, we also have this pattern:

The maximum gap in partner height satisfaction between men and women seems to be around a man being just about 3cm, or about 1 inch, taller than the woman.

For women broadly, this is the least desirable possible pairing, while for men, it would basically be nice if they were about 2 inches or about 5cm taller..., but its not that big a deal to them, they are not that far from their maximum satisfaction.

Meanwhile, this situation is the lowest scoring situation for women.

It is actually worse than the woman being taller than the man.

In case you have not guesed, I am a guy, and I find this ... fairly confusing/interesting.

Basically this means there is a huge mismatch where guys are generally pretty ok with being just a bit taller than their gal, but women find this to be the worst, the lowest possible score they would broadly assign to a partner height difference situation, to such an extent that they'd actually be on average happier if her man was just actually shorter than her.

...

But anyway, yeah, unless I am missing something, it doesn't look like this paper actually answers your question precisely.

What you mention, the uh, height-choosyness tapering off for tall women and very short men does occur to some extent, but we... don't really seem to have that detailed to us, I am not seeing a way to mathematically compare the magnitude.

Also again worth noting, my 'Short King' scenario was kind of a worst case scenario, as it assumes all women would only go for their ideal partner scenario.

Some women do do this, but obviously not all, and some men also do this, but obviously not all... and numbers on the absolute or relative prevalence of that do not seem to be in this paper.