I think you can guess the boundaries of the design answer here.
When you're experimenting with a new technology, everything looks very different and everything looks new. When you look at the net result of people experimenting with it 30 years later, you can see that they were all constrained by the same limitations, and in aggregate - despite many differences - everything takes on a recognizable aesthetic that was produced by the limitations of the day, the technology, the genre, whatever.
These sites are all wildly different and extremely creative, yet they're all recognizably from the same decade or so.
Okay. Now step back from that. Humans - you, me, everyone - have learned a very complex visual language from birth, where certain cues indicate certain things and trigger certain memories. If you show me a picture of a dude wearing bell bottoms, we're probably in the 70s. If it's a girl with a nose ring and a midriff showing it's probably 2005 - but there are smaller cues within that. 99% of people don't think about this consciously when they see an image, or a website, or whatever. They just associate it with a time or a genre.
In design for the web, we've had decades of constraints come and go - we had Flash with its own weird aesthetics, Java embeds, etc. People remember a time and they associate a design style with that time, without really knowing why. So it's a question of choosing from that historical library and constructing something that makes people see what you want them to see. If you want them to associate something with the early 90s, then totally make it look like one of these pages.
The thing I was taught in design school was to attempt to make design that was timeless. This is always kind of impossible - because you're limited technologically in ways you don't even realize at the time, and because you're always influenced by what's around you. But try to make something that will look, in 20 years, like it was new. Very hard to do. Why do it? Because in 20 years, you can always go and make funny vintage pastiche references to what people did 20 years ago. But it's quite hard to make anything that doesn't show its age.
Thank you for the sincere and in-depth response. I hadn't really meant mine to be troll-esque, but I can see how it could come across that way.
My favorite comparison of two sites from the same place is HAProxy: haproxy.com is the one meant for the C-Suite to be impressed, and haproxy.org is the one meant for people who are running it. It's also the reason that I suspect it's not just nostalgia that makes me think of text-heavy sites as being more authoritative (unless it's a share delusion by elder millenials, I guess). If you're going there, you probably are interested in finding a specific answer in technical documentation, not a sales pitch.
Zooming out to other industries entirely, look at cars: it's generally pretty easy to guess the decade of a car if you have any interest in them. However, with some, the designers have occasionally managed to blend the different eras together in a way that I would argue borders on timeless. The canonical example, IMO, is the Porsche 911. Before the 996 model, it had a certain look about it that updated, but didn't change. The 996 had the oft-mocked "fried egg headlights" (plus it was water-cooled, which angered purists), but other than that, it started a more gradual shift in the looks of the car. The 997, 991, and 992 all look remarkably similar from many angles, assuming all are wing-less. That's 21 years of models. Also, the 992 changed its rear vents to a vertical rather than horizontal configuration, which hasn't been used since the 911's predecessor, the 356 - a car that ceased production in 1966. That subtle change, along with some typography for the logo, makes it look way older than it is, but in a good way.
I didn't take your rebuttal / question as trollish in the least! Cars are a good analogy. Designers have a huge palette of historical references and callbacks to choose from, along with newer concepts and whatever is currently in vogue. The difference between a thoughtful design and a bad one is whether those things are blended meaningfully, intentionally, and finally in a pleasing and functional way. So like, not to get flamed if someone happens to be a PT Cruiser fan (because I'm friends with a guy who loves them), but I think that's an example of pastiche. The web design equivalent would be using a blatantly absurd "40s looking" font on the body text as opposed to using it judiciously on a couple of header elements, maybe.
I'm a Datsun guy myself, and I like what Nissan did aesthetically with the new Z, because it's not a replica but it's got just enough callbacks in the form.
I think in all types of design, historical reference is a large part of your "color palette". But you don't want to fall into retro nostlgia or pastiche unless that literally is the whole point of the brand (like, Johnny Rockets or something).
One of the most embarrasing things that can happen to a designer (and I know from experience) is to find out that a style they chose or an image they created because they thought it looked good was inadvertently referencing something they either didn't know about, or which they had absorbed subconsciously. When other people get a reference in your work that you didn't intend, that's a mistake. It's why when I interview graphic designers I'm not looking for whether or not their work "looks good" - which is completely subjective. I'm interested in how versatile they are at mimicking and synthesizing a broad array of visual language, and whether they can explain their choices - including their awareness of cultural callbacks and the intentionality that went into choosing them.
When you're experimenting with a new technology, everything looks very different and everything looks new. When you look at the net result of people experimenting with it 30 years later, you can see that they were all constrained by the same limitations, and in aggregate - despite many differences - everything takes on a recognizable aesthetic that was produced by the limitations of the day, the technology, the genre, whatever.
These sites are all wildly different and extremely creative, yet they're all recognizably from the same decade or so.
Okay. Now step back from that. Humans - you, me, everyone - have learned a very complex visual language from birth, where certain cues indicate certain things and trigger certain memories. If you show me a picture of a dude wearing bell bottoms, we're probably in the 70s. If it's a girl with a nose ring and a midriff showing it's probably 2005 - but there are smaller cues within that. 99% of people don't think about this consciously when they see an image, or a website, or whatever. They just associate it with a time or a genre.
In design for the web, we've had decades of constraints come and go - we had Flash with its own weird aesthetics, Java embeds, etc. People remember a time and they associate a design style with that time, without really knowing why. So it's a question of choosing from that historical library and constructing something that makes people see what you want them to see. If you want them to associate something with the early 90s, then totally make it look like one of these pages.
The thing I was taught in design school was to attempt to make design that was timeless. This is always kind of impossible - because you're limited technologically in ways you don't even realize at the time, and because you're always influenced by what's around you. But try to make something that will look, in 20 years, like it was new. Very hard to do. Why do it? Because in 20 years, you can always go and make funny vintage pastiche references to what people did 20 years ago. But it's quite hard to make anything that doesn't show its age.