I am trying to build a scanner and I have the same problem as described into the Turntable wobble topic. 
Does someone know where I can find the document: table_coupler - extended.STL; 609.95 kb as listed into this topic. When I try to download it, I receive a message that this document is out of date and therefore none accessible.
Thank you.

I had a devil of a time trying to get my platform flat.  Sanding the acrylic cut outs flat and rotating them to try to cancel out any skew bias didn't help.  Pulling the motor and flattening the inside to make sure the motor was square didn't help.

I printed the coupler from the page above and it was a bit better, but still not right.  Then I used both sides of the printed coupler to engage the motor spindle and that seem to be the key.  Assymetric pressure was skewing the coupler and the bed.  I drill with counter sinks for screw holes to attach the bed to the coupler.

Haven't ran it yet, but it is way flatter.  Maybe not perfect.  We'll see.

It looks like most of the wobble was actually coming from the misalignment of FP and FQ. Loosened everything up, inserted a 13/64"-drill's shank with the above-mentioned alignments parts to align everything, then tightened it all up. It looks a lot better now. Let's see how a scan looks now...

I'm now printing a combined FN/FO part with two screws to engage the motor shaft at 90°. There is also an nose/extension that almost touches the motor body to help with alignment.

EDIT: Nope. Too thick.

For those without access to a 3D printer I found a way to mount the turntable tightly to the stepper shaft; Actobotics makes a number of items; for instance they have an aluminum hub for a 5mm shaft which you can attach to the turntable, or even attach the hub to the 4" round baseplate that they sell.  Costs a bit more (in the US about $15) but it is an option if you don't have access to a 3D printer.

Edit:

Here is the extended version of the coupler.  It's 8mm longer, and it should fit the motor shaft very tightly.  You might even have to drill it out slightly, depending on how well calibrated your printer extrusion is.

Anyway, there are also 3 equally spaced holes on the new extension.  These are for installation of 3mm set screw screws.  I prefer to use those on my scanner rather than the 2 original set screws at the top.  But you can use either, or both.

Attach the coupler to the turntable platter, then push the assembly onto the motor shaft until it bottoms out.  There should be only a very small (<1mm) clearance between the bottom of the coupler and the motor face.  And there should be no play at all.

Alternately maybe someone that uses OpenSCAD can modify the original turntable coupler to extends it about 8mm.

I have noticed that out of the many many things it can't do right, Google SketchUp can easily do that by allowing you to select a piece of a model and just stretch it in a direction using the Push/Pull tool... though I don't think it has any way to allow you to stretch it by an exact measurement like 8mm.... it seems to prefer you just drag and draw everything with your mouse.

1.  The hole for the motor shaft is oversized (5.5mm hole vs. 5mm shaft).
2.  THe coupler is too short in length

These two factors contribute to a lot of play, and make it very hard to set the turn table perfectly square.

So for myself, I drew and printed an 8mm extension that has a tighter tolerance (5mm).  I then placed the extension on the motor shaft, and glued it to the bottom of the coupler. 

The new setup, nearly touches the bottom (where the motor face is).  All the play and wobble is now gone.  For those that have similar problems, I'm attaching the STL file of the extension I printed.  It also has 3 holes on the side for adding 3mm set screws.

Alternately maybe someone that uses OpenSCAD can modify the original turntable coupler to extend it about 8mm.