.......another rambler like me sef!! haha.....Some more thoughts: no one that I have "read" about has done any tests involving
A :potency versus plant parts (tops, middle leaves, old leaves, young leaves)
B: time of harvest (summer, winter, fall spring)
C. soil types etc
D. growing conditions (sun, shade, indoors, outdoors...flouros versus halides etc.)
E. drying conditions (natives dry under shade cloth? in sun?...unknowns?)
Potency is already one of the strongest in the plant world......'cept for the nasty plants, like datura or others.
The only personal contribution I can make, is that some younger plants, in spring..absolutley STINK of Salvinorin A....as they were young, I didn't care to kill them to test theory. Just smelling them was kinda scary....

We need some saliva soluable salvia.....that oughta do it.
This paper by Daniel Siebert determined where salvinorin A in all the parts of the plant:
http://aob.oxfordjournals.org/cgi/reprint/93/6/763.pdf?ijkey=55zbQRbeCanlM&keytype=refHowever, this paper did not quantify the salvinorin A. But one of the things it found was that in young plants there are just as many trichomes (which produce and contain salvinorin A) on the leaves than there are when the leaves get older. When the plant matures the trichomes become more spread out. So when they are young the density of trichomes and thus salvinorin A is much higher, which would explain your observation.
Good news though: right now I am in the process of quantifying salvinorin A in all parts of the plant using Liquid Chromatography-Mass Spectrometry (LC-MS). I just finished making my standard curve by "injecting" different amounts of pure salvinorin A (purchased from Daniel Siebert) and logging the values of area of the peak that correlates to salvinorin A. Peak area is the method the LC-MS uses to measure the amount of a substance with a given mass. From this I will make an equation for the line that is produced when I graph amount of salvinorin A injected vs. peak area. So once I start extracting different parts of the plant I can plug the peak area found into the equation to get a value for the concentration of salvinorin A (in mg/g).
The main goal of my research is to identify an enzyme that is involved in producing salvinorin A (it will be a type of enzyme called a terpene synthase). In order to "prove" that I have identified the right enzyme I need to correlate enzyme activity with the production of salvinorin A. So say I find high enzyme activity in the leaves and low to none in the roots then I should find high concentration of salvinorin A in the leaves and low in the roots.
I hope that makes at least some sense, its a little difficult to explain. Once I start actually getting some results I will be sure to post them here.