Braun here from Human Precision. Simple, choice and go/no-go tasks measure different response demands, while devices add their own delay. Here is the practical explanation and the limit that matters when using a browser test.
Different tasks produce different averages
A simple task asks for one response after one signal. A choice task asks the user to identify the signal and choose a response. A go/no-go task also requires withholding responses to distractors. Research consistently treats these as different tasks because added decisions and inhibition usually change response time.
Age is only one factor
Reaction time and trial-to-trial variability can change across adulthood, especially as tasks become more complex. But age does not explain every difference. Preparation interval, attention, practice, stimulus design and response method also matter. A broad age table without matching task conditions can create false precision.
Online averages include hardware
Browser measurements include the person and the device. Touchscreen input, mouse input, trackpads and display systems can produce different timing. Human Precision therefore does not claim that one global average is fair across all hardware.
A better benchmark
Compare your median result with your own previous attempts on the same setup. Look at false starts and consistency, not only the fastest value. The Progress page is designed around personal history rather than a misleading global ranking.
Use averages as context, not judgment
An average describes a group under specific conditions. It is not a boundary between healthy and unhealthy, smart and unintelligent, or safe and unsafe. For a useful browser baseline, take the reaction test under repeatable conditions and return on another day.
Human Precision provides recreational browser measurements. Device latency, practice, attention and environment affect results. Do not use a score for medical, education, employment, driving or safety decisions.