Thanks to Royi Frame's
thread I found the document he refered to.
Interesting to note that they used proper cutting techniques of reduction and thinning to see if there was a difference.
Also, what I have suspected all along, the trees regrew their removed volume quickly and in the case of reduction actually grew back more dense.
Thinning took the longest amount of time to return to pre pruning volumes.
Trees need to be reduced and have 70% to 90% of their foliage removed to make an impact on soil moisture content, then that has to be maintained.
Engineers reports here use the ht of a tree multiplied by a factor to determine the "safe" distance. In heavy clay the factor is 1 and in extreme cases 1.5
For example, 10m tall tree in highly reactive clay soil means 10mx1.5m= 15m away from building. I traced some of the engineers methodology back to a CSIRO document from 1985 called Division Building Research Sheet No. 10-75 Revised Sept 1985. How they get these numbers may be in the last document loaded, look around page 6 and 7.