A review sample and a retail box can contain meaningfully different products. The differences are usually mundane rather than sinister, but they still affect what a review predicts.
Firmware is the largest gap
Press units almost always ship on pre-release software. A patch arrives around launch and changes camera tuning, power limits or scheduler behaviour before the first buyer opens a box.
Reviewers therefore measure a moving target. Battery results and thermal behaviour recorded before that patch may not survive it, in either direction.
Some outlets retest after launch and publish an update. Most do not, and the original numbers keep circulating because that is what search results surface.
Components can be sourced from several suppliers
Manufacturers frequently qualify more than one supplier for displays, memory and storage so that production is not hostage to a single vendor.
Those parts meet the same specification but do not behave identically. Panel colour, storage speed and memory timings vary in ways that careful measurement can detect.
A review unit represents whichever combination happened to be in early production. Later runs may use a different mix entirely.
Early production is watched more closely
The first units off a line receive the most scrutiny, because that is when a manufacturer is still confirming that the process works.
Assembly tolerances, panel selection and quality inspection are all tighter at that stage than during peak volume months later.
This is not a plot to fool reviewers. It is a normal property of manufacturing ramps, and it happens to make early samples flattering.
The box contents can differ
Press kits sometimes include accessories that retail buyers must purchase separately, such as a charger, a case or a cable rated for higher power.
A reviewer testing with a supplied high-output charger may report charging times that a buyer using the in-box cable cannot reach.
Careful reviews state what they charged with. Where they do not, the charging figure is best treated as an upper bound rather than an expectation.
How the gap gets caught
Reader reports are what surface the differences, usually weeks after launch, when enough retail units are in circulation to distinguish a pattern from an unlucky device.
Outlets that revisit products, and communities that pool measurements across many units, do the work that a single sample cannot.
Treating a launch review as an early reading rather than a settled verdict is the practical response, and it costs nothing but patience.