But it is. 1 full second in a 400 is ENORMOUS improvement when at a top level. Take Haile Gebrselassie. He runs 49.97, and regularly closed in 26-28 seconds for the last 200, or 54-56 for the last 400.
Now imagine he runs 48.9, having spent months, maintaining his same endurance work, but doing extra sprint work (I believe he did this before Paris 2003, with work such as 4 x 200 in 24 seconds with full recovery).
If he is able to run 48.9, his speed has improved 2%. This comes into play in many arenas:
1. His normal easy pace can be faster, because he is more efficient and each foot contact is minutely quicker.
2. His workouts get faster, because a 55 second 400 in a workout is a lesser percentage of maximum ability with 48 speed than with 49 speed. A 55 would have normally demanded 90.7% of his maximum capability for 400m.
Now, with his hypothetical 48 speed, he needs only 88.9% of his maximum.
Simply, for a talented, top athlete, these tiny differences in efficiency of stride, relaxation at near top end rhythm, stronger and slightly more elastic muscular and tendon structures, and a more powerful stride and pushoff in the toe-off phase without any additional effort all effects the longer distance running enormously.
In a 200 or a 400, you have very little margin to improve.
The race is so short, there simply isn't enough time to carve chunks off your time. As noted, a 2% improvement form 49.9 is 48.9.
But a 10,000m is 26:30. A 2% improvement on this time is 25:58.
And yet, consider Gebrselassie's nearly meteoric improvement from the beginning of his career to the end.
In Brussels 1994, he won with a PR 27:21.
In 1998, 4 years later, he wins with a WR 26:22.75. His first world record was 26:43. What is the improvement? 21 seconds, almost a second per lap?
How does this happen? There are only 2 ways to do it:
1. Increases specific efficiency at that pace. Through a developed aerobic base and a high lactate threshold velocity, Gebrselassie works at 65 second lap pace until he is efficient, then at 64.5, then at 64, and the times come down as his body adapts and as he nears his maximum ability and his maximal tolerable training load.
2. Increased basic speed and basic stride pattern. If his stride earlier in his career was good for 50.5 in the 400, and his stride rate at this speed was 250 strides per minute, at 200cm per stride, that is his value.
But with years of 60m sprints on the track, short uphill sprints, and weight training (Gebrselassir reportedly uses a great deal of simple plyometric work such as jump roping and soleus/gastrocnemius gym machines. When he stopped using these, he had to drop out of his WR attempt on the 1 hour run in Hengelo because he lost the strength adaptations to his calves by losing this type of training), he improves his basic value.
The strength of his muscles, his hamstring's ability to forcefully push off the ground, and his Achilles tendon's ability to store and transfer elastic energy all improves, even if only marginally. He is faster and more efficient at all paces slower than 49 speed for 400m, and his stride length or rate or both naturally increases.
And you have hit it exactly.
But a top athlete only adds, never replaces, as Renato said was the "secret." Therefore, no one hoping to improve with this would drop previous work used to improve.