I just watched the video, he doesn't say it. 

He suggests that "perhaps" they can reduce DLSS quality (and therefore overhead) to hit the 60fps target. That's the closest thing to suggesting what you think he was, but it was purely speculation about how it could be done given the assumptions made in the video, not a general statement about DLSS or about what will be done. 

He says that dlss works better at lower framerate because it's give it more time to upscale the image.

Part II of guess the base resolution in Witcher 3 NG.

He says that dlss works better at lower framerate because it's give it more time to upscale the image.

Pretty wild. Do you notice a lot of artifacting and blur when you play at this setting? 

This is never said. You can point out the time-stamp where you are confused and I can help you.  

The question was about performance not quality. 

At 4:13 "Would this theoretical performance be enough to run 4k DLSS on Switch 2 to enable 4k gaming in dock mode. To investigate this question we can look on PC and do some fuzzy maths to see where Switch 2 might end up."

He then goes on to compute the hypothetical overhead from doing 4k DLSS on Switch 2 (assuming Orin Ampere @10W): 10.5 ms. 

For a game to run at 60fps, you are targeting 16.67 ms (1sec/60) per frame. 

His point then is that the DLSS overhead takes up about 60% (10.5/16.67) of the frame-time, leaving only 6.17 ms to render the frame. Or in other words, to achieve 60fps at 4k, the Switch would need to run a game at about 162 FPS at 720p. 

Then he suggests that 30fps, is a more attainable goal, because for 720p->4k the overhead only takes 30% of the frame-time. Switch 2 would only need to run the game at 33.3 ms - 10.5 ms = 22.8 ms or roughly 40 FPS at 720p to achieve this. 

The argument was never about image quality, it was about performance. 

He then does go on to talk about image quality between performance mode vs. ultra-performance mode vs. native, but that is an entirely separate discussion from the frame-rate discussion. 

Nowhere in the video is it stated that the image quality (rather than the graphics quality) is better at 30fps than 60fps. Because it isn't. Ultra performance mode is ultra performance mode regardless of the framerate (if anything a higher framerate helps a tiny bit.) Likewise performance mode is performance mode regardless of the framerate. Graphics quality might improve by choosing a 30fps target rather than a 60fps target, but that is for the same reason graphics quality always improves (independently of DLSS) by choosing a lower frame-rate target, you have more resources to render extra effects. 

Prediction time:

Switch 2 mobile:
Max 10W for the whole system, max 7W for the SOC
GPU 3xx MHz
~1 TFP
Heavy 3rd party games will run at 360p DLSS-uped to 1080p


Switch 2 docked:
Max 13W whole system, 12W for SOC
GPU 7xx MHz
1.6 TFP
3rd party games: 540p DLSS-uped to 1440p (maybe the last step from 1440p to 4K using a less intensive method)

Compare 11W and 9W max for Switch 1 when docked and mobile. We will need slightly better cooling this time.
https://www.anandtech.com/show/11181/a-look-at-nintendo-switch-power-consumption/2

Sounds reasonable to me.  Nintendo will likely undercl clock for battery life a will be price focused on other components.  Fair prediction.

If Nintendo were to focus on battery performance, they would not choose 8nm.

It's 8nm The leaker go his info from a credible source.

If Nintendo were to focus on battery performance, they would not choose 8nm.