A time taken by a person with a stopwatch is not merely less precise than an electronic time. It is biased in a predictable direction, which is why it cannot stand as a record.
The bias comes from reaction, not from precision
A timekeeper starts the watch on seeing smoke or hearing the gun and stops it on seeing the athlete cross the line. Both actions involve reaction delay.
The delays do not cancel. Reaction to the start is slower than reaction to the finish, because the finish is anticipated and the start is not.
The net effect is that hand times run fast. Conversion factors exist to adjust them, and the existence of a standard adjustment is itself evidence of systematic bias.
What fully automatic timing changes
Automatic systems are started by the starting device directly, so no human is in the start path. The clock begins at the instant the gun is fired.
The finish is captured by a camera imaging the finish plane continuously, and the time is read from the image rather than judged by an observer.
Both ends of the measurement are therefore instrumented, which removes the reaction component entirely rather than reducing it.
Why records require it explicitly
Record ratification rules specify fully automatic timing for sprint events, along with requirements about the equipment and its certification.
The rule is absolute rather than a preference, because a manually timed performance cannot be compared with an automatically timed one at the precision records use.
Historical records set before automatic timing became standard are handled by separate lists, which is the honest way to represent an incomparable measurement.
Hand timing still has a role
Local meets and training sessions time by hand because automatic systems are expensive and require trained operation.
For those purposes the bias is acceptable, since athletes are comparing against their own hand-timed history rather than against a national list.
Problems arise only when hand times are quoted alongside automatic ones, which happens frequently in recruiting and in reporting.
The transition left a discontinuity
Sprint times from the hand-timing era look faster relative to modern performances than they were, because of the systematic offset rather than any change in the athletes.
Adjustments are applied when historical performances are compared, but the adjustment is an average and does not recover any individual measurement.
This is the general shape of the problem in athletics record-keeping: a change in measurement technology creates a seam that no correction fully closes.

