It could potentially be misleading to link the size of the animals to how much grazing space they need. All grazing animals have a tag that specifies how much they need to eat in order to survive that is independent of their size. Now, I imagine that if you charted size and grazing in vanilla raws they'd look correlational, but my point is that they don't have to be.
Yes you're right in that they don't have to correlate, but in fact they do. Size not only correlates with grazer tag value, but defines it strictly.
It may be not bad as first fough approximation for size to define the grazer value not taking the species of animal in consideration, but the relationship between them is linear in df, and this is wrong.
In fact, animal mass (what is meant by "size" in df) increases as the third power of linear size, and the amount of energy it uses and thus the amount of food it must consume should not generally increase as the third power of linear size, at least for warm-blooded creatures.
If we assume that a creature spends energy on
a) movement and
b) heat dissipation
we see that heat dissipation increases as the square of linear size.(roughly, even slower in fact due to lower temperature of bigger creatures)
Muscular strength grows as the square of linear size, and creature's speed does not grow as fast as it's linear size. So, roughly, power grows slower than the cube of linear size, too.
All in all, grazer tag values as they are are either too small for large creatures or too large for small ones.