Sure, but why would I buy Metric taps? LoL
That is true if your wrenches have a lot of slop (tolerance) but actually for instance the 13mm bolt head is not a true .5" but is actually a .51" and yes a larger 9/16 (.56") will easily fit over a smaller 14mm (.55") bolt head but don't be too surprised if someday that bolt's corners begin to round off, nor should you be too upset if a precision 14mm socket just doesn't fit the old beat up 9/16 bolt without a lot of effort. Agreement that in the case of a 19mm or 3/4" the difference is so negligible no tool should notice.
But no, the 12mm had no real SAE equivalent, so it became a 1/2 inch, the 9 became a 3/8 and the poor 15 (there were a lot of those) became a 5/8. 4mm, 7mm, 9mm, 12mm and 15mms aren't really tolerated around here.
Incidentally for those who work with both and have one of those mixed socket type boxes which tend to mix the different types as you carry the wrench kit around, try my trick of color coding them. Here a red marked wrench or socket is SAE and yellow means it is metric. Since I have over 100 of each and they tend to roll around together in various tool kits I found the color code helps when sorting them.
Getting back to the main topic.
Based on email exchanges with another X2 owner it has been established to my satisfaction that not all X2 or SX2 bases have the exact same height. Someone else measured his own height of the column nut center and where the CAD (from a 3rd mill owner) had 3.25" while I had 3.68", the third person has 3.5".
When I think about it I come up with the conclusion there is more than one mold (or factory?) turning out the mill bases. As long as the top spaces are proper and can be machined properly to fit the table, the height of the base doesn't become an issue, so out there are tall ones, short ones and ones somewhere in between.
So the conclusion is, if you do the addition of the back plate to your column, there is no way around measuring everything first. I would (and do) use the CAD drawing as a reference but the actuality of the measurement's needs to be independently verified for each person's X2 because variations are now known.