Lightning though not AC, shares more in common with AC than DC with its sharp rise times and varying currents. If you look at strikes with a spectrum analyzer you would find that each strike is built of multiple frequencies across the specturm. To do a complete analysis of the effect this pracitce would have on the lightning strike depends on each individual strike.
With that said let's assume that we happen to have a direct building strike to the electrical equipment with its major component frequencies in the 100 kHz range (somewhat typical but again vaires widely) and the that the coil provides 0.001 mHenrys the impedence to the major component of this assumed strike would be (1j + negligible resistance) ohms. If the current portion of the frequencies where a small 5000 amps (strikes can range from a couple thousand amps to 200,000 with super strikes easily exceeding this range) then the voltage drop across this coil is 5 kV. That is significant and will be the cause of loss equipment inside the building in a direct strike.
Someone who is a ham radio operator would quickly be able to estimate the inductance of the coil, but say the inductance is 100 times greater (a mere 0.1 mHenries), now we are looking at 500 kV. I am not going to look up the formula though I should have it memorized somewhere.
I recommend against this practice. NFPA 780 inhibits bends to not exceed 90 degrees from straight and that the radius of the bend be 8" (about 20 cm) minimum.
