um, didnt you just basically describe DLSS ?? just DLSS just uses an AI to make the image less blocky, among other things...
my guess, you havent even seen an N64 on a modern tv, let alone owned one... not to mention you " conveniently skipped over the " thanks to the pass through function" or have no idea what that means.... so doubt you even know what you are talking about
DLSS is to help create super sampling without the need for taxing your GPU, it's not the same as upscaling a retro console's native output. The closest thing to that might be traditional AA, but MSAA and FPGA upscaling are used to achieve different goals.
In modern titles, on PC for instance, you want to run at the native resolution of your monitor or TV when all possible. This is due to pixel layout. The wrong resolutions can look blurry. This wasn't something that CRTs had an issue with, you could run a higher or lower resolution and still get a clear picture just less overall detail in the image itself. For LCD/OLED tech, the exception is running a resolution that's mathematically identical. 1080p can still look good on 4K, since 4K is simply double 1080p and therefore doesn't change pixel layout. This is why plenty of people still use a RetroTINK 5X over the 4K, because its 1080p output still looks good on a 4K screen (though the 1440p looks weird).
An upscaler makes the image mathematically "correct" for the intended source. This isn't actually increasing the internal resolution, it's increasing the external resolution. Unlike DLSS/MSAA, it's not improving the actual visual fidelity or taxing the GPU/system to do this.
When you run legacy inputs through a TV or a reciever, you're getting the original resolution. If you were to put it at the original size, it'd be incredibly small on your TV. Your TV or reciever simply blows the image up to fit the screen, but that ignores it being mathematically correct and can cause it to appear blurry or distorted (and this is ignoring things like interpolation on older systems). It's basically like it's using a magnifying glass. Upscalers increase the image's size, but it's by increasing the output resolution appropriately, rather than blowing up the original without adding anything. I understand this may seem like they're achieving the same goal, but one is fundamentally more compotent than the other.