That is generally correct ... recover everything that produces a smaller-size, narrow response since the gold jewelry isn't going to be larger-size. My back and other body parts I still have definitely limit the amount of bending and stooping I do. Has since an on-the-job back injury in '87, and I was given a cane by my doctor to help me be able to get around back in March of '93. The last five to ten years is really become a struggle ... BUT, I enjoy finding good gold and silver jewelry along with all the coins, trade tokens and other keepers out there. The first suggestion, since you're mainly hunting urban types of Coin Hunting environments and hoping for gold jewelry, is to probably ignore all Iron, ferrous range, signals. Just do not reject anything above the ferrous/non-ferrous break-point.
The 'numbers' will depend on those used by the particular detector you have in-hand. The blanket gold jewelry range suggestion is the 'numbers' can range anywhere from very low foil, just above the ferrous break-point, on up through old and new Pull-Tabs and Pry-tabs and into the Screw Cap range. But that's for most gold jewelry. You can find some better gold at higher VDI read-outs, such as a couple of 24 k, pure gold rings I found that locked-on as a US Penny/Dime, and a few very high k-rated rings or pendants that reached into the modern Zinc Cent VDI range.
You need to find a wider variety of gold jewelry to test sample. Small, thin baby or child's rings, tiny pinky-finger rings and some toe rings will fall in the lower-end of the non-ferrous scale. So will thin gold chins for a wrist, ankle or to hold a neck charm. Then you have a medium range of gold chains to larger, heavier and thicker gold chains. Odd-shaped gold reads differently can due to the shape and orientation in the ground and to the search coil. Chains that are balled up vs chains that are more strung-out. Off-shaped pendants and broaches or religious pieces.
If searching a low-trash locations, such as in a sand-filled volleyball court, etc., the trash is generally much lower than coins and rings and such, so you simply recover all you can. If hunting a popular-use locations, such as around a picnic bowery where there is a lot of common discarded junk, you can 'try' to identify the mosr common trash read-outs and ignore them ..... but any avid detectorist will tell you that doing so, you are very likely to bypass a nice gold jewelry item. So, ignore iron and go after anything out there.
Signal strength isn't really going to matter because that can vary based ups the object's depth, orientation to the search coil, and the search coil presentation. I've found, with over five decades of very avid detecting, that while good gold and silver jewelry can be lost anywhere, my odds greatly increase be hard-hunting the most likely places for it to be worn and lost. Location is definitely the key. I live in a very small town with one park playground and it is very, very seldom used. The playground or the park. A ball diamond that might be made use of for some casual neighborhood' type get-together games six or fewer times a year.
But when I lived in the Portland Oregon metro area, there was an abundance of parks, schools, sports fields and picnic gathering areas that saw very frequent use. There I had ample 'locations' with good jewelry potential, and while there's a lot of costume jewelry out there, my silver jewelry was usually 5 to 6 times what gold jewelry was recovered, and a typical year of detecting rewarded me with ±20 good gold rings plus some other jewelry items. Location is a key, then comes being patient, putting in the time, and making a lot of recoveries.
Monte