Bigtime400, you're mixing up concepts. "Discrimination" and "target separation", and "see through ability" (aka "target averaging ability") are three different things. So for example: a machine may reject a nail (or foil or tab, or whatever you've set it up at) excellently, all day long. However, if a coin were *under* that nail, or close beside that nail, you might get a masking effect, and miss the coin. Do you see how the machine did not lack the ability to pass the nail, so therefore, you can say it was "good at discrimination"? The problem comes in when you start talking about masking.
Then there are machines that do a great knife-point job at target separation and discrimination both at the same time (albeit you have to interpret some clicks and clacks). Those are typically your 2-filter detectors, like the Whites Classic, shadow X2, various Tesoro's, etc... The trouble with those is that ... while they are great for getting targets close to disc'd out items, the trade-off is, that they are not power-house deep seekers.
The 3rd category of "seeing through" varies from machine to machine. Ie.: can you put a nail OVER (not next to) a coin, and still get a good signal? Note: of course the TID will be skewed, but can you at least discern that there is a conductive target there, or has it gone totally silent now? Usually the 2-filter machines are also good at this test too, but it varies from machine to machine, and this 3rd point entails a lot of operator interpretation, in addition to the machine.
The usual rule of thumb is that the deeper-seeking a machine is (Explorer, CZ6, etc...) the worse they are at target separation and masking (but can be good at "discrimination" though!) The deeper seeking machines are simply "seeing" more ground in each pass, so it's inherent that they are seeing more targets at a time, than a knife-blade less sensitive machine is seeing. So there is less to sort out on the one machine, verses the other. Speed of processing also plays into it.
Personally, I think the Explorer, given it's great depth on coins, is a darn good compromise on all fronts. I even use it (with small 5" coil anyhow), for iron-riddled demolition site hunting. But sure, if a site is very nasty, I grab a shadow X2 or even my Compass 77b.
You might also check out the Minelab Musketeer. It has nearly the depth (?) of the Sov and Exp, but has better ability in target separation and masking. It lacks the other bells and whistles (like the expanded tone ID of the Exp, etc...), so that's the trade-off. A friend of mine had the Musketeer, and we tested it in a variety of situations. As long as you had the disc. set low, so as just to reject iron (ie.: relicky mindset hunting), the Musketeer effortlessly would get signals on coins next to, and sometimes even under a nail (depending on the positioning, size of nail, etc...). And if we rejected foil or tabs, and then put a penny or dime under the foil or tab, it would get a signal no problem. That is, it averages the targets together, for a net-positive signal, well.