Detail just gets in the way of imagination, I hope they're grateful

Yeah it's more because devs seem to have this weird obsession they have to have compatibility with what seems to be an absurd amount of different sex toys

The problem before PBR was that it was really hard to make textures uniform and look correct under different lighting conditions. That's why piss-filter existed so much during that era. Slapping a color filter over everything to unify it was a way to solve that issue.

You can actually invent modern fidelity from first premises if you just work through the problem of creating a game world. For example: GTA III. That game had 32mb of VRAM available to portray an open world city. Rockstar invented a mesh * texture streaming system that loaded in pieces of the world right around the player, and more distant objects based on size. That way distant large buildings would still show up rather than pop-in. Meshes and textures were kept to 2KB or less file sizes. Files were duplicated on the CD to limit seek times. Roads had to be kept shorter and car speeds had to be kept in check so the player couldn't out-run the streaming. Fog and fading LODs were used to hide pop-in.

If we start there, keep the same fidelity and then remove the memory limit of 32MB, we can now load in a much larger radius around the player, perhaps even the whole map. If we had 8GB, we almost certainly could. If you're trying to make a cool immersive city, you may then start thinking of increasing its size and density. At the same fidelity as GTA III, you could make a pretty large city. But the problem you run into is that it would be a mostly empty city with many blocks of nothing to do except watch the pedestrian and traffic sim play out. People still have to author the gameplay part. So you either get the repetitive side task/collectable method (the Ubisoft way) or it takes much longer to dev because you're adding hand-crafted missions that cover all locations. In fact, this was kind of the problem with San Andreas. While a lot of people (myself included) like the game, it did have an emptiness problem and if you remember they did add collectables to help fill it out.

If you make the new city just a little bit bigger than the old one, but increase it's fidelity (therefore increasing immersion due to the added visuals, sounds, animations, details, etc) then you basically get to the same place without the problem of a big empty world. The world size is a bigger and more dense, but it's manageable. You can reroute the memory gains to other things that make the world seem more alive. That's how we get to GTA IV from GTA III. It's not an ideological commitment to photo-realism at the expense of everything else, it's just an optimization problem between fidelity and the time it takes to author something.

That kind of problem solving holds for all types of games, stylized or realistic, open world or discrete levels. When you have a memory increase and you've done with it all you want to gameplay-wise, there's nothing left but to add some more detail to everything. I only say this because I've thought through this exact same problem step-by-step only to realize I've just re-invented a modern game.

I guess the real problem is that these games are products. A game that holds Xbox 360/PS3 fidelity would have to offer something more than what the original games offered. The game looks the same, has the same gameplay, is the same size/scope. But everyone making the game is paying 2026 prices for everything so the game has to justify 2026 dollars so the owners can make 2026 returns on their investment. Of course a game made as a passion project, not subject to market demands, could be made anyway the devs wanted.

As a game developer or artist, it would be very boring to keep making the same game over and over again, so think there is a genuine desire to increase scope. It's easy to say "focus on gameplay rather than graphics" in that regard, but making gameplay is difficult. Unlike graphics, there isn't a clear fidelity or quantifiable depth. How do you improve on your favorite 360 game? Introduce quality of life features, fix bugs, and then what? If you start trying to modify systems to be more deeply authored or simulated, you end up with different gameplay anyway. The artists and level designers care about their part of the process as much as game designers. They too want to go above and beyond what they've done before. It's kind of unfair to say that those people should just stop worrying about doing more with their work and let games only be driven by the gameplay people.

That desire to do more is how you end up with modern games, even without a market force. An artist is going to go "I tried making wheels with only 8-sided cylinders, but I have the memory budget for 16-sided cylinders and some bevels. I'm going to do that." Again, it's not them trying to endless pursue photo-realism, just desiring a better approximation of form. Of course there are neat technical solutions like using a 8-sided cylinder but then using textures as caps so the wheels look round. But that comes from a limitation. If doing with geometry works, it'll be done with geometry. That's why games don't do that anymore except for ones chasing a certain aesthetic.

halo 3 still looks amazing

If the game is immersive/enjoyable it really doesn't matter. I don't look at pixels or polygons when I'm playing a game but at virtual objects. It doesn't need to be true to the real world. True to the real world might even be considered a boring creative choice.

The picture is also just not high res enough to tell

Starfox 64, right?

Anything that is not textbased or ASCII is bougie.