I have just wasted a lot of time making a discovery about scanning linearisation targets. I was trying to scan a paper which had a coating of GO on an i1io2 table. It looks like the GO is affecting the glider on the arm. Bottom line is that you can't get repeatable readings on GO coated paper with an i1io2.
I switched to hand scanning with the i1Pro2 and got repeatable measurements. I also scanned a 2000 patch 3 page HPR Baryta set twice on the i1io2 with perfectly acceptable results. One patch out of 2000 had a dE of 0.99.
Hopefully, this will save others going down the same rabbit hole.