You can think anything you want, but your analogy is ridiculous and that doesn't bode well for your initial assessment.
CPU production variance
IS that high, however chips that perform that poorly are
thrown out because they won't perform as advertised. This is called binning. Even within the SoCs that go to market some of them are faster than others (see: overlocking, undervolting, binning).
In terms of Intel CPUs the only difference between a top end part and a low end part is this binning. One CPU is unable to hit high frequencies at reasonable voltages and is thus called a lower end part and targeted to lower frequencies. Similarly a dual-core CPU is (often, not in all cases) a quad-core CPU with botched cores which are disabled.
In the display manufacturing world the same thing occurs (you'd be amazed how many panels are just thrown away), however there's no market for "lower binned" screens because you can't simply "turn them down" like you do with CPUs and sell them as lower end. This has two results:
-Huge batches of screens are thrown away.
-The tolerances for what's considered an "acceptable" screen are fairly wide.
This means that a single sample is NOT representative of the retail product, especially if this sample comes from the manufacturer. This is equally true for Apple and Samsung (and everyone else).
With any display panel of the quality expected for mass-production consumer devices you are likely to see variances like:
-White point balance (+/- 1000)
-Peak brightness (+/- ~40 cd/m2)
-Brightness uniformity (+/- 10%) <--- generous as you sometimes see up to 20% variance here in many consumer grade displays
-White point uniformity (+/- 10%) <--- this is the famous Apple "yellow blotches" effect.
-Backlight bleed (LCD-specific obviously)
The final result in terms of reviews is that reviewing one unit is insufficient to make conclusions. A reasonable display review would look at a dozen retail samples (at least) and assess the total possible variance in quality. Manufacturers don't want this to happen, which means the reviewers would have to obtain these units and the whole thing becomes too expensive to pursue.
I'll even cite a review of an Apple rMBP (supposedly famous for their high level of quality) which shows a relatively poor variance in brightness (one area of the screen is 20% brighter than another):
http://www.notebookcheck.net/Apple-MacBook-Pro-Retina-15-Late-2013-Notebook-Review.120330.0.html
-----
Obviously the reviews aren't completely meaningless (when you see big differences between two displays that does mean something), but you guys are arguing about tiny variances in dE between two high end screens as if that's not simply within manufacturing tolerances. It makes the entire debate pointless and clearly a case of "fanboy vs fanboy."
OLED's subjective superiority as a display tech is unquestionable simply due to its plasma-like properties (as the OP's title suggests). In fact I would say that for media viewing even the Note 3's display is far above the current top-tier LCD (even though it doesn't measure nearly as well). That has very little to do with precise calibration values.
