Of course, given the outrageously generous depth of field with wide-angles, you really could 'range focus' the old fashioned way and suffer nothing. - David Lyga
Unless you've used a pure wide-angle lens it's hard to appreciate the differences, there's less of the exaggerated pulling effects at the edges of a similar SLR lens, it's less distorted.
I use the equivalent of a 21mm Leica lens on my 5x4 camera and in most of the images you wouldn't realise how wide angle a lens I'd used. In the early 1990's I used a 21mm on my Leica for a few months and was extremely impressed with the lens.
What Ian Grant wrote about Biogon and Distagon in the Hassy is a good way to look at it.
In RF land, it was not until 1935 when Michail Russinov (or Roosinov) pioneered the use of aberration vignetting (to improve corner illumination) and designed vignetting filter deposited on the inner surface of the wide angle objective exterior lens..
Also, M.M Russinov pioneered the idea of enlarging the size of the entrance pupil as angular coverage increases..
By 1946 M.M.Russinov lens design (Russar) was patented and documented widely.
Zeiss and Leitz came late to the game and that explains why they used workarounds.
When Ludwig Bertele was commissioned by Zeiss in 1951 to design a wide angle lens for Contax and Hasselblad, Bertele could not obtain master patent for the use of single meniscus at each end of the lens as Russinov had already covered this.
At the International Congress at Stockholm in 1956 Ludwig Bertele paid tribute to Russinov for his highly interesting and original solution to the problem of securing adequate illumination in the picture corners etc. etc..
In 2011 Dr. Hubert Nasse, senior scientist at Zeiss and chief optical designer wrote in Camera Lens News 41 published by Zeiss Carl Zeiss AG Camera Lens Division:
"In 1946 the first patent for a new kind of symmetrical wide-angle lens was applied for by the Russian lens designer Michail Roossinov. It looked as if two retrofocus lenses had been combined with the rear elements together and thus had a symmetrical arrangement of positive refractive powers close to the aperture, surrounded at the front and back by strongly negative menisci.
As of 1951, Ludwig Bertele carried this idea further and designed the legendary Biogon on behalf of Zeiss..."
For me, the Achilles' heel of retrofocus wides is geometric distortion. By definition a retrofocus (reverse telephoto) lens is not symmetrical, which introduces geometric distortion, which is complicated and expensive to correct. As a result while many RF wides exhibit minimal and/or virtually invisible amounts of barrel distortion, a partially corrected but still very visible amount of simple barrel or complex (mustache) distortion is usually considered a good performance even in a $2,000+ SLR wide. There are a few exceptions around, but generally it seems to me most new SLR prime wide designs are pretty lackluster when it comes to geometric distortion correction. My guess is that is because most people are shooting digital and simple geometric distortion can be largely corrected post-capture. Unfortunately while there are some very sharp SLR wides, in my experience the correlation between price and distortion correction is relatively weak.
No complaints about the 2.8/24 and 2.8/28 Ai-S's; great '80's lenses. But then, maybe they're not real wide angles...
The Nikkor 24 f2.8 was a very good lens for the time, and one I still like. I imagine shooting it side by side with a Super Angulon would show up some significant differences.
And, of course, that doesn't mean your pictures would be better with a better lens, but that wasn't the question here.
I am not a lens design expert but would agree with the theory that wherever there is some additional constraints in lens design, either price or lens quality has to suffer, or both.
I have a small and quite well made Voigtländer 15mm on my Voigtländer Bessa-L and the pair is light, portable, and fairly inexpensive.
A 15mm on a SLR would undoubtedly be more expensive, heavier, and very likely sacrifice some image quality on the altar of design complexity. I very often take pictures of architectural subjects so distortion correction is something I deeply like.
I would say that, overall, wide-angle photography is something that would better be left at range-finders. If you are the kind of persons that goes around with two or three bodies, the wide-angle body should be a range-finder (or quasi range-finder, like a Contax G2). It's not just a matter of distortion correction but also a matter of weight and size.
Enjoyed the lens design story, thanks georg16nik.
As several have noted... In terms of ease of focus and composition... rangefinders are well suited to wide angle, while SLR's are better suited to telephoto.
In the late 70s a "teacher" at the traveling Nikon School said wide angles don't have distortion but exaggerated perspective. He followed with a sensible proof using the ratio of distance to foreground objects and distance to background objects. Whether or not this was provided to explain away true distortion in WA Nikkors versus WAs by Leitz I don't know but 1) It made perfect sense, and 2) I've never had much problem with "distortion" if I framed the image correctly. Usually, I was after the FG/BG thing so I coaxed it as much as I could. And, with the Nikkor Retros I could actually afford the 28s, 24s and 20s that let me get the images I was after. If Leitz was the only source of wide angles I doubt I would have ever picked up a camera. I have no doubt the geometries and optics can affect things here but must also think at least some of the spiel behind $3k 35s is to explain away Leica's antiquated production workflow.