I test in a different way.
First I have to have been in a color temperature environment for a minimum of 30 minutes for my brain to bring its white balance in sync with the ambient CCT.
When I wake up in the dark (more than 30 minutes asleep causes my brain to lower its white balance below 4000k), I see 4000k and above as increasingly blueish
When I wake up in the dark, I see 3000-3500k as neutral white, and 2000k as only slightly warm white.
When I have been sitting under 3000k incandescent for a couple of hours, the incandescent does not look warm (yellowish/orange), instead if I shine 4000k or higher on the wall next to the 3000k, the 4000k+ look increasingly blueish
During daylight hours, after my brain has spent a couple hours sitting by a window, if I use an LED with CCT 4000k or lower, the LEDs looks warmer (yellowish orange), to increasing degrees.
When adapted to daylight my 3000k E21a and N219c look warm (yellow/orangeish)
When adapted to daylight my sw45k looks white with strong pink tint
When adapted to daylight my 6500k optisolis looks white with slight green tint
The pattern is that my brain thinks a spread of more than 1000k between ambient adapted brain, and turning on an LED of a different CCT than ambient, produces awareness that the LED is either cooler, or warmer, than ambient adapted brain white balance
everything depends on what the ambient CCT is, and that the brain has had at least 30 minutes to adapt its white balance to ambient. Then a light source of a different CCT will be perceived as warmer or cooler.
I have tested this extensively by putting a light on my nightstand, and then using the same light during the day. At night 4000k+ looks white with a blue cast, during the day the 4000k or less, looks warm (yellowish orange cast)
