If, for instance, the ones on #2 and #3 are more than 500 ohms different from each other, the lower-resistance one gets starved for spark (that would be #2, here...).
How do you justify that? I was taught:
2 and 3 are part of one coil's secondary loop. Once current flows in the loop, it is the same throughout the loop based on the total resistance that the loop contains and the voltage achieved to create the spark channel conductor.
When the coil's energy field collapses, the voltage rises in the secondary until a spark channel is formed across both gaps. Since there is no current flowing during the voltage build up, the voltage would be the same at each spark plug, and resistance would not be a factor until current is flowing.
Are you saying the once the spark event begins, the voltage drop causes one plug to shorten it's spark event due to lower voltage? If so, how could the remaining spark plug sustain an arc without current flowing through the spark plug that no longer has a conductive path through it, as they both are part of the conductive path of the coil secondary circuit?
The caps should be 7500 ohms if they are OEM caps (10,000 ohms on post-1975 bikes), with +800 ohms more being their burned-out limit. Replace with 5,000 ohm caps today.
I've never found anything less than 10K caps on the stock 550's. And, no Honda documentation that specifies anything other 10K Ω for the plug caps.
5K ohm caps do increase coil heating, and slightly increase the coils power consumption, as they will discharge father toward depletion. The higher currents will also wear the spark plug's electrodes faster.
I don't understand why 5K caps would be "better", besides being the only ones readily available.