Because the compass works correctly on east/ west magnetic headings (or on the magnetic equator).1
Suggestion: learn more about how the "dip" of the earth's magnetic field, combined with the fact that aircraft normally bank to turn, is the fundamental cause of the tendency for the compass to "lag" while turning if the aircraft's current heading contains a northerly component, and to "lead" while turning if the aircraft's current heading contains a southerly component, in the northern (magnetic) hemisphere.
The key factor is not whether the aircraft is turning toward the south or the north, but rather whether the aircraft's current heading contains a northerly or southerly component, and how much. For example, when flying SSW, the compass will "lead" regardless of whether the aircraft is turning left toward a heading of S, or turning right toward a heading of SW. Conversely, when flying NNE the compass will "lag" regardless of whether the aircraft is turning left toward a heading of N, or turning right toward a heading of NE. So as an aircraft's current heading approaches due (magnetic) W or E, the compass error drops toward zero, regardless of whether the aircraft is turning from a more northerly heading (i.e. turning toward the south), or is turning from a more southerly heading (i.e. turning toward the north).2
Understanding the key role played by the aircraft's current instantaneous heading, rather than the direction of turn, is probably the real key to clearing up the OP's confusion. It seems the OP believes the "UNOS" rule is based on the direction of turn, which is not the case--the key factor is the aircraft's current instantaneous heading.
But ultimately a pilot is usually going to be modulating the turn rate by the artificial horizon (via the bank angle) and/or the turn "coordinator" or the turn rate indicator, in which case the behavior of the compass is not of great interest until the aircraft is getting close to the target heading, when the compass will be used as a guide as to when to roll out of the turn. Hence the focus of the test questions on how the compass will be behaving as the aircraft approaches the target heading.
New suggested knowledge test question for OP-- if you are turning to the right (toward the north) from magnetic 180 to magnetic 225, according to the correct application of the "UNOS" rule, should you "undershoot" by rolling out "early" (e.g. do you expect the compass to "lag"?) Or should you "overshoot" by rolling out "late" (e.g. do you expect the compass to "lead"?) If you recognize that the key factor is the target heading, not the direction of turn, you'll know the answer. 225 is heading with a southerly magnetic component, so you should expect the compass to "overshoot" as you approach this heading, and so you should continue the turn until the compass has gone slightly past the target heading.
Secret knowledge: there are some specific cases where the compass's tendency to "lead" while turning when the heading contains a southerly component (in the northern magnetic hemisphere) can actually be used to a pilot's advantage. It tends to give an "early warning" of the first beginnings of a bank, which can be helpful. For more, see this related ASE answer.
Within limits. There is also the rotational inertia of the compass card itself to consider. If you are in a steep turn, with the compass behaving erratically and often "freezing" or even reversing its direction of spin, you certainly would not expect the compass to read correctly the instant your actual heading passes through due east or due west. That's why we keep the bank angle shallow when we are relying on the magnetic compass. And that's one reason a heading with a large southerly component affords a much larger margin for error as per (the non-standard technique mentioned in) the last paragraph above. The critical thing is that the aircraft's heading not be allowed to gain any significant northerly component, where the compass will "lag" rather than "lead", and the odds of that are best if the pilot chooses a target heading with a large southerly component.
Same as footnote 1.
Related ASE answer of interest, with some useful links -- Why do we undershoot/overshoot with a compass turn? -- but beware, in places the answer emphasizes the direction of the turn rather than whether the target heading is a northerly or southerly heading. This may perpetuate the misunderstanding that has apparently ensnared the OP.