There are 3 places the 0.006 could be coming from: Knee, Saddle, and Table. Or a combination of all 3 which is the most likely. It is not likely that the table is worn that much, or even that the table ways are worn that much. Consider that the table weighs 100's of pounds and once cranked to it's extremes is is putting a lot of twisting force on everything. When unlocked, my knee ''twists'' about 0.003, measured about 3 inches off to the side of the knee, as the table moves through it's travel. I can see this movement on the DRO reading. My machine is about the same size as yours.
Doing the same test again, start at zero in the center and look at the change as you move the towards the ends. My guess is that the indicator will move positive as you work toward the ends. Then repeat this test with the knee and saddle locked as
@brino suggests. You should see less change in the values. Then doing the test after unlocking the saddle and knee one at a time should tell you where the real problem is at.
Once the machine is cleaned up and lubed, then start adjusting gibs on each axis. I'll bet most of that goes away.
0.020 and 0.030 backlash is what I would consider in the range of normal. You always have to approach your work from the same direction to compensate for the backlash, this is standard practice on any machine. The nuts may be adjustable, but don't tighten too much or you will get tight spots.