Hi, 2-speed;
The diode will clamp the negative EMF spikes, but not the positive. When I worked my way thru Electonics Engineering school, we played with this same thing on Hondas, as they were hard on their points. What we found was the diodes caused increased pitting of the points, building up the material on the moving point very quickly. We also discovered several other things that would probably translate to cars:
1. Required condensor capacity is directly proportional to coil current. A 4-amp coil, for example, needed about .25uF in the condensor to balance the point wear. A 2-amp current draw needed about .15uf, or about 50% as much. The highest-current test we ran was at 5 amps, which required a .33uF capacitor. If you had too little capacitance, the pitting caused the moving electrode to grow and the ground side to pit. Vice versa, too.
2. When we tried using diodes, the spark from the coil had an unbalanced output: the high-voltage swung wildly over the RPM range and was always below the voltage produced with a capacitor (condensor).
But, that was in the 60s, and you know what they say - "If you remember the 60s, YOU WEREN'T THERE!"