If you have to selectively choose areas for it to stand out, doing so would not be representative of the whole experience, and would only confirm that the overall improvement in the experience definitely would not be worth the performance impact.
I'm confused by this post. The RTX in Metro does indirect lighting. Showing scenes with no indirect lighting in a vs comparison video is pointless. It's like you want to compare shadow quality and use scenes without any shadows. That's not a valid criticism?
Then the fast paced chase scene from Cyberpunk, also not a good base for any comparison. I bet no one could see differences in any other visual effects like texture quality, AO or shadows judging by that scene.
The WD Legion benchmark is again not representative of gameplay, although there the reflections are somewhat visible and easy to spot.
I'd argue that if we would take these exact scenes and take any other visual effect to compare like texture quality low vs ultra, SSR quality low vs ultra or any other effect... no one could spot them as easy as with the raytracing in this shitty compressed youtube video. I'd argue that in direct comparisons, raytraced reflections and lighting are way easier to spot than differences in most other visual settings.
If raytracing effects are worth the performance hit is very subjective and also depends on the game and what exactly it raytraces and how. For me personally, it's worth it if I manage to get at least around 50fps at 1440p. Which my 3080 easily manages.
Might be a different story for you of course.