That's fair enough. I can definitely see how the PS4 Pro would be a disappointment if performance was the main reason for upgrading. The performance modes seemed to rarely hit a stable 60fps, and usually quality mode was the better option. Plus a lot of games (looking at you Bloodborne) never got upgrades. I never owned a base PS4, only a Pro, and was quite happy with it. 

Cyberpunk: Phantom Liberty DonkeyKB and 60fps, 1080p Tears of the Kingdom are way above PS4 ability and that's just the start. In 2 years time Switch 2 titles will be stunning.

What was funny is I popped my pro on a LED TV and base ps4 on an OLED...  concluded the base + oled looked better than the pro + LED.  Before that time I never knew how important panel quality was.  

For me it's hard to feel strongly about numbers or performance once we get to a certain threshold. There's an intersection between art and hardware which means simply putting TOTK at 1440p60fps does not make it something which "feels" above PS4 level hardware.

I played it for a short while and was like this looks cute but it looks like a Switch game (lack of good AA doesn't help).

I'm looking forward to Nintendo showing us what they do when a game is built for the ground up for Switch 2, maybe Kirby may be our first glimpse of that tomorrow

Of course very different games & scope but some of the combat environments Ni No Kuni 2 on PS4 reflect something which you won't catch in TOTK. I'm very much an effects, lighting, art guy and you just can't get that from a resolution and FPs bumps. Of course there's a nice middle ground to be found

On the PSP discussion, a big part of the reason many games were worse than their PS2 counterparts is the control scheme, rather than the system's power. The lack of a 2nd analogue stick limited options for camera movement, and the 1 stick it did have was nowhere near as good as what the consoles had, nor were the shoulder buttons. Lower budgets for handheld titles and development often being given to B-teams rather than say Insomniac also didn't help, but something like GTA being less fun than the PS2 games has more to do with the controller limitations than the processor limitations.  X-Play's entire review of Ratchet and Clank Size Matters was just Adam complaining about the PSP's control setup.

Yeah TOTK is pretty solidly a late 7th Gen to 8th Generation game (PBR, moderate volumetrics, decent AO implementation, etc.) 

PS4 Pro probably could run it similarly to Switch 2, maybe with even better image quality if using check-boarding to get a good looking greater than 1440p output (though Switch 2 could've used DLSS if they had put in the work to get motion vector support.) 

Definitely wouldn't use it as a standard of what to expect from a Switch 2-only 3D Zelda or Switch 2 games in general.

This seems like a very good and interesting study, albeit based on one game only. As an economists and analyst, I want to interpret the data and the three attached graphs, namely regarding the following...

Figure 3: QoE
Jump from 7 to 15: Very large benefit and statistically significant at 95% confidence interval.
Jump from 15 to 30: Very large benefit and statistically significant at 95% confidence interval.
Jump from 30 to 45: Still large benefit and statistically significant at 95% confidence interval.
Jump from 45 to 60: Fairly large benefit and statistically barely significant at 95% confidence interval.
Jump from 60 to 75: Small benefit and statistically barely significant at 95% confidence interval.
Jump from 75 to 90: Small benefit but statistically NOT significant at 95% confidence interval.
No benefit beyond 90.

Figure 4: Score
Jump from 7 to 15: Very large benefit and statistically significant at 95% confidence interval.
Jump from 15 to 30: Large benefit and statistically significant at 95% confidence interval.
Jump from 30 to 45: Small benefit but NOT statistically significant at 95% confidence interval.
No benefit beyond 45.

Figure 5: Mouse Movements
Jump from 7 to 15: Very large benefit and statistically significant at 95% confidence interval.
Jump from 15 to 30: Large benefit and statistically significant at 95% confidence interval.
Jump from 30 to 45: Small benefit but NOT statistically significant at 95% confidence interval.
Jump from 45 to 120: Small benefit but NOT statistically significant at 95% confidence interval.
No statistically significant benefit beyond 30
No benefit beyond 120.

Interpretation
1) The benefits up to 30 fps are clear and universal. The quality of experience increases for everyone, the mouse movements become smoother and most people score higher points.
2) Quality of experience improves for almost everyone up to 45 fps, mouse movements feel smoother for the majority of people, but only a minority can score higher compared to 30 fps.
3) Quality of experience keeps increasing up to 90 fps, albeit at smaller and rapidly diminishing rates. Score does not improve at all beyond 45 fps. Mouse movements continue to feel smoother beyond 30 fps, albeit for a rapidly shrinking population and at increasingly smaller rates.

In short, 45 fps seems to be a dividing line and 60 fps is a good reference point for quality and consistency. Anything beyond 60 is niche at best. Better yet, anything beyond 90 is just placebo for 99% of the populace. 

So 60 to 90 is statistical significant....  huh, so there is benefit passed 60 fps...  I'd act surprised but that was my position.  

My position that 120 offers benefits over 60 fps still stands.  Has to stand because 75 offers benefits over 60 fps.  

I never said that it was statistically insignificant... 

I also never said that there is no benefit to 90 vs 60... 

That was not your position, your positions were:

'Diminishing returns past 120 fps. Sure. But objectively the average gamer does in fact benefit above 60 fps.'

and

'I'm astounded people think 60 fps to 120 fps is diminishing returns.'

It's not against the rules to change your position, in fact I think that's healthy when presented with new info. But claiming a W when in reality your initial position has been proved largely incorrect is a bit lame...