A couple of other related links that might be of interest, which take the concept of silent tuning somewhat further in a couple of other contexts:

Fully automated silent tuning: https://athena.westpoint.edu/items/9acd01cb-3e48-467c-ba5d-72afd8df90fa (Melville & Hamilton)

Using an algorithm to go straight to the correct tuner settings, no need to squint at graphs in the field: https://github.com/conniest/InstantTuning (Stillinger & Melville)

If you’re selling them, yes, it’s important to measure antennas in the wide open. If you’re USING them, on the other hand, especially for non-hobby things where success is a requirement, you should be measuring them in situ, where their pattern and impedance are distorted by the same factors that will affect their actual use.

Mostly not a ham topic, I know, but…

I didn’t say you were an idiot, but you are pretty sadly uniformed. You made the assumptions, not me. And by the way, there is zero difference between an antenna analyzer and a VNA used in S11 mode. Antenna analyzers (I have three different ones spanning 30 years) can can plot real, imaginary, or both. VNA’s ( I have two of them) can plot real, imaginary, or both. If you think there is an actual functional difference, I’d like to hear it.

Yes, the one you want is a Vector Network Analyzer, but buying one of the old ones without knowing what you’re doing is going to be a problem. Even the right cables for them can cost hundreds of dollars, even the genderless connectors with repeatable SWR per connection cost hundreds. Good luck if you buy a unit without them.

You can find T1 Network analyzers which look awesome and are completely useless and sell for a couple dollars.

One was made by a cottage manufacturer, and I picked it up at a ham swap meet. It wants the dipole to be about 10% long and has a variable capacitor driven by a motor that goes in the center.

And of course the SteppIR are a variable length vertical in a fixed length radome.

There’s a manufacturer at Hamvention who sells a box that implements a physically variable length dipole. It uses wire and rope. I think there’s a German company that makes a remote balanced tuner. In the kind that I first saw from SGC for sailboat use, and was later cloned by MFJ switches both resistors and capacitors in, I think, a pi matching network.

I’m amazed how you instantly decide I must be an idiot. Let me try again.
I believe that nowadays basically any VNA worth having is going to be usable for everything we’ve been discussing. I also believe that just because your antenna analyzer might have had nice features doesn’t mean that a “random”, typical, common, average, normal example of an antenna analyzer was always like that. I think a normal example would be closer to a mfj-259d.
I also think most people’s plots were typically a rectangle where the x-axis represented frequency and the y-axis represented another real number. So since a complex number needs two real numbers to express, but you only have one axis left, you would need two different plots – one for the real portion and one for the imaginary portion. You can overlay them if that’s supported, but it may well not be. I used one not even ten years ago (though it may have been a few years old at that point). It could only plot one variable at a time, so if you wanted complex numbers you needed to read them off as single points because you couldn’t get both real and imaginary portions onscreen at once any other way. And if they’re not onscreen at the same time, then like I mentioned with the wind, they can have varied by the time you take the other reading. At least the single-point measurement has both parts taken simultaneously. Data-logging to a computer could solve it, but we’re talking portable use here.
There’s a real UI advantage in plotting differently. And there’s a real difference in the liklihood of any random device supporting various features if it’s calling itself a VNA versus an antenna analyzer.

“That’s not the point and you know that, I assume.”

Writing respectively, I did not see that as your point.