As far as I'm aware coning and quartering is by far the most used and accepted technique for sampling soils.
They only way to verify the homogenization would be to analyze several samples from different quarters and check if all of them are statisticaly equal to the average.
Alan, I'm aware of the wide spread use of cone and quartering technique for powdered samples, due to its inexpensiveness. But, I couldn't find any publication to justify its acceptance for soil samples. On the contrary, I found a paper recommending the riffle splitting as most efficient than cone and quartering and grinding and sieving methods (Schumacher et al. (1990) - Comparison of Three Methods for Soil Homogenization). In addition, if you are working with contaminated soil, as I am, none of them seem to be suitable from safety perspective. I'm currently using a "twin shell V-blender", adapted for small mass samples.
Looks like you have more research than I on the subject actually. I had the impression you weren't pretty sure about it, my knowledge comes rather from following stabilished protocols used on my lab than from research, so I don't think I can be much more helpful. Anyway, good luck on your project!