The "threaded rod" that adjusts a rolling mill for an exact thickness output is very large. For something that will roll steel, it probably should be 2". I think they are at least 1" on my smaller jewelry rolling mills.
The McDonald type mill shown in the plans here just uses the foot pressure to space the rollers, as precision thickness isn't needed. This allows the billet to be slid all the way in and then drawn outward under pressure. The foot powered cam applies far more force than one would suspect. I originally thought that it wouldn't be sufficient, but after seeing the videos of them and talking to people who used them, it turns out that it is plenty. I am sure a foot valve and an air cylinder could do the same job, but that may be unnecessary since the foot powered arm works great, and is a fail-safe device. If your foot stops pressing down, the roller drops open.
Go ahead and try a set of beveled rollers if you want. They can always be expensive door stops when you find it won't work. The thing you are leaving out of the equation is that the bar will not be coming out straight at you. It will be coming out in an arc toward the thicker side of the bar ( spine). That would make you have to move sideways with it as you draw the bar, and thus couldn't apply the foot pressure anymore. If you were adding more than a few thousandth of bevel, it will come out at a significant arc. We adjust the roller screws off-parallel to do this deliberately in rolling sheet stock to make curved jewelry items, but that is a far different thing and scale. In a 24" bar of steel/damascus, it would be a big deal.
Just to do a little math:
If the bar is 24" long and .250" thick, and you beveled one side to .225"(10%) to make a very slight bevel, that would make the chord 24" and the arc 26.4". That would make the segment height almost 5" ... which means that is how much the bar would curve for only a very slight bevel. If the bar was 1/4" at the spine and you wanted to bevel to 1/8" at the edge, that would make the bar come out as a semi-circle arc or 180°. These are imperial, and may be a tad more than the actual change through a mill, but you can see how much curvature would be formed by a very slight angle in the roller alignment. Additionally, there can be a lot of tearing along grain boundaries when applying uneven pressure to the bar when rolled. This can make for all sorts of trouble far worse than a curved bar. If the edge cooled faster than the spine due to the difference in pressure, it could tear the bar edge.
Beveling a bar has to be done at forging heat with the pressure from edge to spine, not longitudinally. You simply use a set of beveling dies in a press to do that.
