It's all too typical for nVidia, in my experience. They've now walked it back, because of the tremendous bad publicity it earned them, but the damage is done. It's just too bad AMD doesn't have a big stockpile of graphics cards to sell, because a lot of people would have switched on the spot.
As for the general question with real-time ray-tracing, I think it's a long way off from usability. Less than 5% of gamers spend $500 or more on a graphics card, and those are the only cards that yield barely playable frame rates when RTRT is in use. I'm utterly unimpressed by DLSS, so that wouldn't mitigate the situation at all for me. Why increase the visual quality with RTRT only to degrade it with a messy and artifact-laden AI upscaling algorithm?
Many of the effects have a noticeable impact on visual quality, but not enough to justify the performance cost.
It's really kind of a chicken and egg problem. Games are going to get better looking without RTRT over time regardless, so rasterization performance has to increase. RTRT performance isn't going to go up without more silicon dedicated to specialized processing, which eats into the budget for increasing rasterization performance. And none of this is going to work at all except on very high end cards, which less than 5% of the market will purchase and use. The software needs to come along significantly to justify the hardware changes needed, but the software is only going to trickle in until there's more hardware available that's capable of running it, and at much lower price points.
I don't see the new consoles making much of a dent in this, because the only RTRT they'll be able to do is fairly light.
Maybe things will creep along slowly, gaining momentum, but it seems more likely to me that we need some kind of tipping point, with no obvious path to reach it.
Maybe that's why nVidia is trying to bribe and/or bully the tech press into pushing the RTRT dogma. It's not going to work.
