RAZZ is right, UT engine uses the FOV ( in degree ). So it is quite correct already.
As for the zoomable scope, currently the viewable area increases with a lower power and decreases with a higher power. This is totally wrong.
OK, let me tell the truth. I will consider this problem in 2 ways.
First, if you put your eye directly onto the eyepiece, without an eye-relief, which means 100% of your FOV is inside the scope, you will actually see it works in the opposite way as in INF. When at 3x power, you can see a black circle around the viewable area; as you raise the power, the black circle expands while things look bigger; the actual seen FOV is constant ( let's take 10-degree for example ). You keep raising the power gradually, and the black circle will gradually go outwards. When at about 6x power, you can't see the black circle anymore, because it has gone outside the FOV which the eyepiece occupies and/or your own eye's full FOV. For an even larger power, due to the fact that the full scene ( 10-degree FOV in our example ) has gone outside the full viewable area as mentioned above, the actual seen FOV becomes smaller than 10-degree.
Second, let's consider a real aim. When aim with a scope in real life, there IS a certain amount of eye-relief -- you don't put your eye directly onto the eyepiece. Usually you keep a small distance about as big as your fist, from the eyepiece to your aiming eye. In this situation, you can't see that "black circle effect" because the FOV which the eyepiece occupies is much smaller than the previous example. What you see will become a thin black rim ( assuming your scope is black-colored ) around the maginified area ( simply put, it's the eyepiece, haha ). In this case, no matter what zoom ( power ) you use, your viewable area is -- simple -- the eyepiece, always.
But I was talking strictly from in game and I don't think the team is modeling muzzle rise with torque, my intention was simply to point out that the muzzle rise would overcome any such drift downward. 
