There should be a difference, although it may be less than what you'd expect. Suggest trying again and being very careful and methodical. Place the coil well away from anything metal, use a plastic or wooden ruler and hold the end perpendicular against the bottom center of the coil. Wave the target coin steadily back and forth right past the ruler while staring straight down at the ruler. Make a mental note of the depth you get consistent audio. I always take note of two different depths: 75% or more proper discrimination tone ID (for instance the high squeal of a coin on the MXT) and secondly the max depth of any kind of all metal signal. On my MXT as well as my old V3i and VX3, the all-metal "channel" would get a about two extra inches.
I've tested all those same coils on my MXT and V3i (except for the Ultimate 13 which I didn't get until after I sold my last VX3). The ultimate is deepest giving me a good repeatable coin tone on a DIME down to 10.5" and all-metal down to 12". Next is the 10x12 SEF: 10" with good tone ID and 11.5" all-metal. Then the D2: 9.5" in disc and 11" in all-metal. Finally the 950 concentric: 9" in disc and 10.5 in all-metal. Both my V3i and VX3 would air test virtually identically to my MXT.
Now, as far as air test vs in-the-ground field performance, I've found that the more mineralized the ground, the more difference there is between air vs in-the-ground. In my mild Boise soil, there's maybe a 10% reduction in depth compared to air. In other words, the 10x12 SEF will air test a dime with good tone ID down to about 10" but only go about 9" in actual ground. All the other coils are similar, seeing about a 10-15% reduction in depth when in actual ground. When I head up into the mountains and into some heavily mineralized ground, sometimes the dropoff is as bad as 50% or so. That's why most people say air tests don't matter, because mineralization plays such a big role.
However, I do believe air testing is extremely useful in a controlled environment to compare different detectors and different coils to one another. Air testing represents the maximum possible depth in 0% mineralization. So you can change coils out on the same detector and get a very good idea of relative performance of the two coils against each other. There are other variables that may be in play, like halo-effect, which gives a little more depth in the ground than you'd otherwise expect, and EMI, but still air testing is still very useful for making comparisons in a fairly controlled way.
One thing I noticed about the MXT vs the V3i/VX3: the MXT's reliable tone ID depth drops off less quickly in mineralized ground than the V3i/VX3. The more mineralized the ground the less able the V3i/VX3 is in identifying the target compared to the MXT. Even in my mild soil, the MXT ID's deeper and more accurately than either of my V3i's or the VX3 could. I think this is due to the V-series finicky and frankly inferior ground balance system. But yes, you need to run the V3i hot to get good depth. The MXT at the preset level was as deep as the V3i when it was run hot. The V3i was at least a little quieter, but no deeper, and a helluva lot harder to get set up right.
Hope this helps. Good luck!
