The book does not say, and I found no develop commentary; speculation follows:
It does make some sense that the faster pace would result in less than the expected number of miles and the slower pace would result in more than the expected number (walking faster is harder, and walking slower is easier). So that's one small potential reason for the difference.
Also, as user Mathaddict pointed out in their answer, it is likely so that all the numbers are divisible by 6, making them far more useful for hexagonal grids.
And that's about all I can think of... The only other thing to I would note is that these are all approximations anyway, and not very good ones at that:
Here is the DnD Beyond version of the table:

The "Distance per Hour" entries are all rounded down from what we would calculate using the "Distance per Minute" entries (4.54, 3.41, and 2.27 miles).
Furthermore, if we take the 400 foot pace and extend it to 8 hours, we would get 192,000 feet, or 36.363636 miles.
Similarly, with the normal pace of 300 feet, we'd actually get 27.272727 miles.
And with the slow pace of 200 feet, we'd get 18.181818 miles.
This isn't any sort of reason for why the numbers are wrong, I'm just pointing out that even if they were "right" going off the "Distance per Hour" entries, they would still be wrong going off the "Distance per Minute" entries. More accurate Distance per Day entries would likely be 36/27/18