You forgot to mention that those studies are NOT DONE WITH LCD'S or AMOLED displays. They're done with high contrast printed material under perfect lighting with people with similar good eyesight. Tha'ts one hell of a difference since no LCD or AMOLED display is comparable to that at all.
So, it is not clear at all that we'd see much of difference consistently over 400 if you could actually get the same device with both a 400ppi and 450ppi of the same type to test with and perform blind testing.
The only place something like that could show would be in black on white text, mostly seen in reading books, some web pages and sometimes the UI.
For the high contrast to exist, the screen should be pristine, right out of the factory, the angle of view straight on with no reflection off the screen and a high level of light from the display (so, observed in good light, but not outside).
As soon as any of these conditions are not met, you cannot have maximum contrast and it is very unlikely you could see the difference even with eagle eyes.
Ideally, compared phones should have identical light coming off the display (checked with a light meter), the same gamut and be calibrated the same, the same display technology and the same display construction (laminated versus not laminated). Often, it is preference for other aspects of a display which influenced how on display resolution is perceived as better. If they know one display has a higher resolution than the other, this alone can influence perception regardless of the actual performance of the display.
All that put together means that there may be a few situations with a select group of people were 450 is very slightly for them. But, even for them, it would be a small part of their overall user experience.
Real life is not the lab, and screens are not printed pages. Still a lot easier to read text off a printed page than a lcd screen.