It's clear _to us_ because when we think about such things, we think in
Newtonian terms. I'm not at all sure it would have been clear to people
in the middle ages; when you throw a ball, it whizzes by so fast, it's
hard to be sure how it's actually moving.
If they asked an archer to fire an arrow through a distant window, he'd
aim slightly above it. You can't spend dozens of hours every week
shooting arrows at targets without learning to compensate for gravity.
The theory of impetus went through a number of variations over the
millennia. Despite the unsourced diagrams on the Wikipedia article (see
the Talk page for more details) the usual medieval view of impetus was in
the context of ballistics: an arrow or other projectile was fired up at
an arrow, it traveled mostly in a straight line, then slowly curved away
as the impetus was lost and gravity took hold, and then finally dropped
straight down.
http://en.wikipedia.org/wiki/Theory_of_impetus
While it isn't a good model for arrows and cannon balls, it's actually
not too far off the real-world case of a light projectile in the face of
air resistance.
We can be sure that Aristotle was not a juggler, or spent much time
watching jugglers. If he was, he never would have come up with the
impetus theory in the first place.