Since posting I came to a similar conclusion that 1. the print orientation could add some error if 2 different parts are printed 90deg to each other. 2. That my holes in the motor and idles ends are loose where the 10mm smooth rods go (i'll get a measure and insert here).
So with the combination of these 2 points would easily attribute to a 0.3 to 0.4mm error. I think with these Igus DryLin Bearings I'm making my life a little harder as well.
So yes printing all 3 parts in the same orientation as we printed the carriage would be a good test. The smooth rod holes for me on all parts of the printer are way too big and on Y and Z I'll have to put tape on them. For my own design I try to go for having to run a drill down the holes to 'clear them out' but I do understand people making them loose to cover the variety of printer performances out there.
The funny side for me is we are discussing such a small error on a sub $1000 printer built on a dining room table... love it.