Basically, I'm pointing out how the ground can change a reading significantly enough to make in-air or on top of the ground readings rather meaningless. Sure, a lot of people are blessed with lightly mineralised soil where they live but, for example, the places I hunt can knock a SILVER quarter air VDI down to near iron.. within 3-4". A couple of weeks I dug a target (which turned out to be a silver quarter) that gave a mostly iron reading, that sometimes jumped just a bit higher, that was only a few inches down. I can't recall how much silver I've dug up in the general area (but a lot), most at pretty shallow depths, and this area get's hit hundreds of times every summer by quite a few detectorists. I often watch them, and on a few occasions BS with them, and see how many toss coins down to check the ID or wave targets across the coil. Come the fall, I'll hunt these same areas and pick up a load of goodies. This is just one area, there are many more that exhibit the same properties.
As for my statement ("I don't know how many times I've hunted behind somebody that 'knows' where the goodies are on his/her meter; I find the goodies instead") and 'paintsticks' being misleading.. nope. It holds true whether the targets are glued onto a paintstick or just held in the hand. Getting a solid repeatable signal generally depends on one's disc. settings ( and GB, and sens., and..) and if the detector is set 'wrong' for the ground over which a person intends to hunt well, there may be problems obtaining that nice solid signal. An example would be having the disc. set to knock out anything below pulltabs when the ground mineralisation 'pulls' the VDI down enough to drop silver/copper readings into the foil/nickel range (sometimes lower) when buried just a couple of inches.
As for the 'debate' regarding air and ground tests, I don't think that there's much of a debate insofar as it's pretty well established that, especially when mineralisation increases and with a VLF detector, target ID accuracy degrades more significantly at increasing distances into the ground matrix vs. the equivalent distances in the air. Sometimes, just laying an object on the ground may be enough to skew the ID readings.
Regardless, I'm basically pointing out the limited utility of air VID readings vs. real world (in-ground) readings. I basically do a version of what you're talking about (sans the paintsticks) by checking the on-ground VDI of targets vs. freshly buried and noticing the discrepancy between the 2. We can probably argue back and forth ad nauseam (hope I spelled that correctly), but I see your point and mine, I believe, is sufficiently clear, to make much more debate an exercise in juvenile nitpicking. Anyway, that's about it.. I type way to slow. ..Willy.