What moves the threshold
The discrimination threshold is not a fixed quantity. Longer notes make the difference easier to catch, while shortening them makes it harder. Timbres rich in harmonics can supply additional cues, and background noise pushes performance down. Even for one listener the value changes across the register. Trained musicians are reported to detect smaller differences than untrained listeners, but that gap is not an innate constant; it reflects familiarity with the task of comparing notes as much as anything else. Comparing figures measured under different conditions therefore tells you nothing.
A different axis from hearing
Pitch discrimination and hearing are easy to conflate, yet they measure different things. Hearing asks how faint a sound you can detect, and clinicians assess it frequency by frequency in a sound-treated room with calibrated equipment. Pitch discrimination asks how fine a difference you notice between two notes that are both perfectly audible, with loudness held constant. Someone with unremarkable hearing may or may not discriminate pitch well, and the reverse holds too. If the way you hear is itself a concern, a pitch discrimination measurement offers no evidence bearing on it.
Why it suits measurement
Because pitch discrimination rests entirely on a relative comparison, it lends itself to measurement. Play two notes at matched loudness and timbre and vary only the pitch, and the judgement of which is higher barely depends on the frequency response of the speakers or the volume setting. Attempts to measure absolute sensitivity instead yield results determined by the hardware rather than the ear. Keeping the notes within the range of ordinary speech ensures built-in speakers reproduce them, avoiding the situation where an unheard note cannot be attributed to either the listener or the device.