The quote in post #6 by its wording is claiming the "frame" is built within 0.1 mm. If this same claim applies to all the different sizes of frames then the measurements of the _frame_ should be within this measurement for the various axes. To build a frame with this degree of precision is commendable but then spoil it with badly positioned or mis-aligned good quality parts is foolish. Measure with a micrometer, mark with chalk and cut with an axe. Yep, good plan - not!

It would be interesting for two, three or four folks with the same makers model to compare their frame measurements to verify this claim.

The specs for the accuracy of a machines mechanical movement of say 0.03 mm is the minimum amount of movement the machine can make reliably between two points on the advertised axis.

The repeatability is the axis ability to return to a point of origin after having been moved a set distance to and from that origin. This is often expressed as a value (say 0.015 mm) over a certain distance (Note rounding). The "certain distance" is rarely published by a machine builder because of the advertising wars of competition so in order to make true comparisons one needs to look at the linear positioning system by a given manufacturer and look up their specs for the components used.

Take HiWin for example, their profile rails are manufactured in at least 3 different grades, each grade specifies the tolerances that can be measured over various distances. The same applies to the motion components, ball screws or R+P mechanics. The accuracy and repeatability of these positioning components needs to be combined with the other components used in the mechanical assembly. Example, a 5:1 ratio reduction gear set and/or belt drive set will be different to that of a 10:1 reduction set. These values can all be calculated and some can be fairly easily measured with a DTI or digital calipers if they have sufficient accuracy (and repeatability) for such measurement.