Ed in SoDak, you hit the nail right on the head. As an example: I once had a CZ6 user tell me he could tell a coin under a rusted bottle cap (as if .. his CZ could see through it). To illistrate this, he puts a rusted bottle cap over a penny, and swings. Sure enough, a signal. When I asked him "but can you tell that signal apart from bottle caps with no coins underneath it?" He says "of course, it'll sound different". Imagine his surprise when I tried this test on him: While his back was turned, I put out 3 rusted bottle caps, with a coin hidden under only one. Then I challenged him to swing over them, and tell me which bottle cap was hiding the coin. Believe me, he never made that claim again, and he began to have a hefty respect for the concept of "masking" after that
There is also the issue of "averaging" (to change the subject a bit). Some detectors are better at this than others. That's where, for example, you have a dime which, on the Whites scale, reads 79, right? So you put a tab, which reads 30, over it, and you will end up (depending on the machine, the distance between the targets, etc...) presumably with a reading of around 54 (the average between the two). But I've noticed that phenomenom only on certain machines. Others simply only read the object that is on top, especially if there's any appreciable distance between the targets. And this is more pronounced with 2 conductors, not iron vs conductors.