The radios will perform well as long as you do not overload the first transistor amplifier (the LNA). That means run lots of current through that transistor. Also, do not share GROUNDS with the motors or power drivers, even inadvertently; thus avoid Efields and Hfields coupling into Radio VDD or Radio Ground. Another trick is to use LOW Data Rates, so motor transients are short compared to a radio data-symbol duration.
The shielding effectiveness of cable strongly depends on the percent-coverage of the cable-shield; thus a foil-shield should capture more of the electric field.
For magnetic shielding, you need the magnetic fluxes to CANCEL; that means the stepper-motor current must have NO OTHER PATH than the cable. Since there will be parasitic capacitances inside your motors, there will be currents/charges flowing through wiring-insulation to the CASE of the motor. You need to capture that current. I suggest you evaluate attaching the CASE to the cable shield, while electrically isolating the CASE; in a plastic drone, that may be simple.
EDIT Standard PCB foil is 35 microns thick (0.035 millimeters, or 1.4 mils) and weighs 1 ounce/foot^2. Frequencies at 4MHz (125 nanosecond Trise, Tfall) will be somewhat attenuated; frequencies at 16MHz (30nanosecond Trise, Tfall) should be attenuated by 2 Nepers (2 * 8.6dB); that atten of 17.2dB is not much to protect your RF Receiver front end circuits from the overloading magnetic flux of 1 amp being switched in 30 nanoseconds. PI matching circuits should be chosen to provide High Pass Filter behavior, not Low Pass Filter behavior. That means the components both before and after the PI matching are contributing to reduction of induced Magnetic field overload energy of that first RX amplifier. Thus capacitors into the PI and Out of the PI are guaranteeing High Pass Filtering; if you see a DC path from antenna to the LNA base or gate, you have failed.
One of you questions is "the open connections soldered to the board". Yes, those matter, because the distance and area from RTN to HOT wire are the loop area creating a flux that will overload the radios. Make that area as small as possible. Something like this is good:
simulate this circuit – Schematic created using CircuitLab