I can't speak for George, but on the set up I am buying, I helped in it's initial design and manufacturing. On the ball screws, there is only a thrust retention on 1 end of the ball screw. Thrust and secured from 1 end only, the other end is supported but can float in the axial moment. This was intentional on our soon to be My machine. The reason was the screw is steel, the frame is Aluminum. They expand at different rates, so allowed for the difference in lengths to take place in the end shaft support. This is done on all the axis screws. The total movement between a hot day and a cool night is not really huge, but it is still there.
The motor to the shaft is coupled through one of those special flexi couplings just like what Hardinge use on their super precision lathes, but only smaller ones. The other option was to use wavey spring packs on the end of the leadscrews to keep tension on the screw the whole time. But there has not been any problem with what we have made so far. The only upgrade is going to be more powerful stepper on the 3 axis to handle the ruffing out of steel parts.
Neil