I would just like to add some numbers for consideration, based on my calculations. I know that theory and practice may differ in important details, but certainly the practice cannot violate the laws of physics (only I can screw up the calculations). I do run my car using BHR ignition and a dwell table based on those calculations, and cannot really complain.
I designed my tables based on the fact that the D585 coils self-trigger at approx 8.4 Amps. I use dwell times so as to limit the maximum current to 7.5 Amps to have a safe margin and remain in control of the ignition. So if you only take peak currents, I'm looking at 30 Amps peak, but only for a rather short amount of time.
The current flows through the coil only during the dwell time and increases exponentially with a time constant determined on the coils resistance and inductance. The necessary dwell time depends also on voltage you can supply.
The point is, at 14 Volts, I can reach the 7.5 Amps in 4.15 ms of dwell time. At 9000 rpms, 4.15 ms corresponds to 62% duty cycle (at 1000 rpms, it would be 7%), so even if the peak current of 7.5 Amps was flowing through the coil for the whole dwell time, it would still be only 4.67 Amps on average.
If you actually take into account that the current increases from zero to the peak along the exponential (http://hyperphysics.phy-astr.gsu.edu...mgele/indt.gif
), the average current at 14V and 9000 rpms with a dwell of 4.15 ms is 1.4 Amps (and 0.15 Amps at 1000 rpms). I have limited my dwell times to 2 Amps maximum average current, which is only seen at lower voltages (below 12 Volts) and towards lower rpms. The highest I have for normal voltages (such as 12.75 Volts) is 1.79 Amps with 4.79ms dwell at 9000 rpms.
That's for one coil of course. For 4 coils, the average current they draw at 14 Volts and 9000 rpms is about 5.2 Amps, even though for a very short time before firing, they peak at 30 Amps.
BTW, that only holds if you use custom, more aggressive dwell tables. TeamRX8's tables had a max average current of 1.9 Amps at 14 Volts with 4.71ms dwell at 9000 rpms (peak 8.15 Amps), and Oltmann's had 0.73 Amps with 3.20ms dwell at 9000 rpms (peak 6.25 Amps, but 7.51 Amps at 1000 rpms). MazdaManiac's dwell tables were at 0.31 Amps with 2.32ms dwell at 9000 rpms (peak 4.87 Amps, but 9.80! Amps at 1000 rpms which would mean the coils self-trigger).
I don't have all the info for stock coils (apart from there being a dwell time of 1.54 ms at 14 Volts and 9000 rpms), because I don't have a model of the coil (resistance and inductance). For the D585, I approximated the model (the time constant) based on some oscilloscope snapshots I found on the web.
In the end, even if you size the wires for maximum current, I think that the average current reflects the kind of load the wires are subjected to from the thermal perspective. Could this be the explanation why the stock wires apparently work fine over the years?