and here's why:
#1.. In the 'real world' you're going to encounter ground that is 'different' from your coin garden ground.
#2.. In the real world, you're going to encounter a variety of coins that are still in excellent physical condition because they were located in 'kind' ground. Thus, they have retained their size and shape and quality of alloy make-up/
#3.. In the real world, unfortunately, you're also going to encounter some really ugly condition coins and other targets because they were located in 'bad' ground. I don't refer to more iron mineralized grodund, but also ground that contains salts, or ground that has been artificially treated with fertilizer and other chemicals. In some cases, overhanging trees and/or adjacent shrubbery can add to a chemical change in the ground that will have an impact on the deterioration of various coin/artifact metal alloys.
#4..In the real world, a wide array of masking trash, both ferrous and non-ferrous, can be associated within the electromagnetic field of the detector/coil combination. In addition, the trash could be a single item or multiple, and in various sizes and shapes. Also, it can be positioned in a co-planer arrangement with the desired target, or perhaps shallower or deeper. In line with the axis of the desired target or off to the side.
#5.. In the real world, coins/artifacts might be located in firm ground and positioned such that they have been there, unmoved and in compacted soil, for a year, many years, decades, or longer!
#6.. In the real world, desired targets can be associated with changed/changing ground environments, such as a coastal beach after a good storm, in some renovation work, in a plowed field.
#7.. In the real world, coins and other desired targets are not always going to be positioned 'flat' or in the most favorable position top be 'detected' and/or produce a decent audio signal and/or visual Target ID.
#8.. Planted targets are immediately associated with an environment that has surround undisturbed ground, and a small patch of disturbed ground due to the 'planting' process. This, in itself, can have an effect of altering the otherwise anticipated response.
I am sure I could think of a few other things to add, but in short, there is no way to "get it right" when designing a planted test garden.
Now, do not take this as me saying that planting coins to evaluate performance is a bad thing. It isn't. matter of fact, it is an excellent way to learn what sort of things you might encounter in the real world because you've now added some ground mineral signal to the mix compared with waving some target in the aid in front of a coil.
So, my suggestion is that, while there isn't a "right way" you might want to consider planting a variety of targets at various angles/positions, and at depths from 2" to about 8".
use this to learn a little about your detector(s) and coil choices, but also learn a lot more from in-the-field recoveries. never rely 100% on visual or audio Target ID systems.
Happy Hunting,
Monte