OK, don't freak-out, it's easy to deal with...
First, look at the picture of the locking paddles and notice the bottom right one. That paddle often interferes with the base hold-down bolt below it and when you crank backward in the Y direction, the paddle hits the bolt and tightens the gib. Unscrew the gib lock to remove the paddle and grind the paddle down short enough so it won't interfere anymore. Then put it back in.
The weight of the table is supported by the bed and there is no adjustment there. So, if your bed is see-sawing, there's nothing you can do -but they get it right at the factory. Mine is 2 years old and it holds under 0.0002 from end to end. It's rock solid from the zero to 10" mark and shoots-up 0.0002 at the extreme ends (between 10 and 12") where you wouldn't normally mount anything. That measurement can only be checked by putting a TDI on the spindle and running the table back and forth. The same is true for both X and Y direction.
Now for your problem. Everything hereon describes the X direction but it's the same thing for Y...
Next, look at the pic of the gib position screw with my finger pointing at it. There is one at the other side too and before turning one in, the other must be backed-out out first. The gib is long and tapered and it adjusts the table rotation -meaning, if you crank out all the way in the X direction, it goes in a perfectly straight line instead of going off at a slight angle. (hope that makes sense).
Now, look at the picture with the TDI. It's fixed at the bed and indicating off the table. Clean that surface and run the table back/forth. The needle shouldn't move (much). If it's off by more than what you feel is acceptable, you need to push the gib to the left or right and capture it into position using the gib position screws. Give one side a twist out and the other a twist in, to make your adjustment then, snug the paddles then, back them off a little. Check the travel a couple times. When you first crank the wheel, the TDI will show a little deflection because things will shift a little due to the tiny amount of tolerance in the fit. What you're looking for is stability in the needle over most of the travel. Keep adjusting the gib one way or the other until you're satisfied.
Don't go nuts with it... The table is 24" long and if you have 0.001" deflection from end to end, you have a non-linearity of 0.0024 degree. -Good luck noticing it. Machining to a tenth of a degree is hard. Hundredth of a degree... -Very, very difficult. Thousandth of a degree? Only in your wildest dreams...
Let me know if you want to know how to do the top side...
Good luck...
Ray
PS: I know what you're going to ask me... How close is my X-Y table linearity set? Close as possible which is about 0.0005 from 0 to 12" in either direction. Once you get this set, it doesn't really ever change -that I'm aware of.