So I've had a bit more of a think and a play with the numbers. It seems that theory and practice are in harmony. If we assume water is added as 30% of the weight of the oils then this equates to 23% of the weight of the raw soap (ignoring lye, fragrance and colours which would make it proportionally smaller again). Thus, if all the water were to evaporate then we would see a weight reduction of a little less than 23% - just as Linda reports!
Ta dah! QED, as my old mathematics teacher used to shout. It seems
ALL the water evaporates during the 8 week cure. This would be borne out by the lack of change over time beyond 8 weeks.
This would give us an easy way to check when we could assume the weight is stable. Divide the weight of the non water ingredients by the number of bars (whole + wastage) and compare that to the weight of a cut bar. If they are close, or even the same, then they can be assumed to be stable. Of course, this assumes the bars are the same volume (size and shape). It helps me because I've never had the variation issues that Terry has been having with the commercial soap cutters.
Many thanks to Linda for helping me out on this. It might seem a small thing to most folk here but I like order and predictability. My wife says I have "CDO" which is a severe version of "OCD" where the letters are in the correct order.

I hope this helps someone else too.
So, now to figure out what "maturing" process is happening beyond 8 weeks for Castille soaps to remove the sliminess.
