I didn't read the whole thread so I'm sorry if I miss some good points.
At the risk of earning some disdain, I'm going to declare I think it's slightly overstated to grant it the claim of "greatest of all time".
Let's look at what it supposedly has going for it, and here's why I think that.
Price? It was $700 and that's almost $900 in today's money. I think it might have been particularly good for the enthusiast class, but I think something accessible to only a tiny fraction of the market makes for a limited argument of "greatest of all time".
Longevity? How is longevity defined?
I don't think simply looking at "can it at least play 1080p at high" as a fair measure since if you were only after that in 2016/2017, you could simply buy a GTX 1060 instead for much less money. A GTX 1080 Ti was more of a 1440p or even fringe 4K product, no? It has definitely not retained that ability. I have no problem calling it relevant enough still. I have a problem with moving goal posts and say its longevity hasn't slipped like the rest of the market. It's definitely needed concessions the same as the rest, and people are also still using lesser GPUs than it still today. Do they win longevity points? No? Why not?
There's also a particular point that i think needs said here. What if that longevity is earned simply because the majority of the rest of the market has slowed in growth? After all, an RTX 4090 being much faster doesn't mean much when the majority of the market isn't seeing similar levels of uplift, and the majority of the games have to be developed with the majority of the market in mind (I believe this is also why there's so many complaints of poor optimization of games in recent years, because games are rubbing up against a slowing market growth). When the majority of the games have to be playable on this more slowly increasing average point, then of course the GTX 1080 Ti will remain more relevant. Let's not forget that the GTX 1060 sat as the most popular GPU on Steam's hardware survey forever, and then the GTX 1650 displaced it! Growth has been very slow until the huge over supply of RTX 30 series finally flowed out and the RTX 3060 displaced it... which I believe still only equals the GTX 1080 Ti, give or take? New generations are two years (and growing?) now, not one. Nobody batted an eye at a graphics card being relevant three generations later back then. Even mid-range ones often did that. If just doing that garners a stake to such a prestigious claim, then I say that's more a sad sign of how much things have slowed overall and less of a sign that a particular GPU from the tail end before that slowing point is the best of all time. We saw this with CPUs already; when gains were slowing in the early 2010s, a lot of CPUs saw some serious longevity. Does every single one of those deserve awards for a stagnating market around them? I'd say hardly.
Strong uplift over predecessor? It was a cut down Titan (I think?) but it launched as the refresh of the flagship in name though so let's look at it like that. Some other generation to generation uplifts were very big too (some bigger). The GeForce 6800 comes to mind (though similar to the Core 2 being so good, it's simply because what it replaced was so bad). The GeForce 8800 series comes to mind as well.
None of this is to say it's bad. It was fair at most of these things despite my counterpoints, but in my mind there's a pretty obvious outlier nVidia has put out as its truly "greatest mistake" and it's not even close. I'm talking about the GeForce 8800 GT of course. It offered something like 95%+ performance of the flagship 8800 GTX a year later, at a third (or a bit more) of the price. Furthermore, this lowered price was accessible to the majority of the market. nVidia practically obsoleted the entire market (including the rest of its own product stack) with that single product. It was often heralded as "the only graphics card that matters" in reviews, and it just about was. There was almost never a reason to buy anything else. The GTX 1080 Ti is a far cry from any such claims.
And yes, the 8800 GT didn't have the same longevity in year count, but again, generations went from one year to two. Resolution was also growing back then as we went from CRTs to 5:4 LCD to widescreen (which was 1200p or even 1600p at the upper end) in such a short time, which no doubt accelerated how fast GPUs of that time aged. If GPUs actually got fast enough to make 4K native playable, then you'd be looking at something closer to similar surrounding market states, and the GTX 1080 Ti would have faltered in such a case long ago. But instead it gets the benefit of recency bias because of a faltering market state where instead we're developing upscaling techniques. Instead, we're being given massively cut down chips and being told 1080p is all we need because it's what most of us currently use (sounds a lot like Intel's "four cores is good enough" all over again, huh?). If ever we needed another "greatest of all time", another 8800 GT, now is the time. Another GTX 1080 Ti right now would just be an RTX 4090 Ti, something costing 20% more but being 25% faster. Or, or... imagine something 5% slower but only 33% as expensive. I know which one I'd rather have, and it's not even a question. Now imagine that latter scenario actually happening, and then being considered aged a few generations later on top of that! That is how good the GPU market used to be. Put the GTX 1080 Ti in that market, and well... it's probably just another GPU.
......
