Oftentimes the 'guy in NZ who . . . etc . . .' is right and his words of support are much appreciated. But replacing a capacitor won't get anyone round a defective coil problem.
Sometimes, a coil with a small break inside will keep working for a good while as the spark will jump it until it gets too big, and a replacement condenser (of any sort) will have the effect of bringing things back to life if the original has started to fail. But it is unwise not to try to establish the state of the HT winding. Some originals are still , rather surprisingly, OK - but the twin risks of an internal break (which can usually be established by simple resistance measurements) and of failing insulation (which can't), should never be overlooked. Being perfectly honest, I'd say 8 out of 10 original coils I see fail to match the Lucas specs on test, often showing problems when heated to the Lucas test temperature of 50°C.
By contrast, 90% of 'modern' rewinds test good for continuity and insulation, although there are variations in performance depending on the exact recipe used by the particular winder. The basic Lucas performance spec is that you want a K2F to produce 90%+ of all sparks, across 5.5mm test gaps with three points, at 130rpm of the magneto. A good one will produce them all at 110rpm, and exceptionally down to 100rpm. A good MO1, as per B series machines etc, should do slightly better than a K series mag.
Many of the modern condensers fitted don't do as well as the originals and are the the reason for premature failure of rebuilt instruments. The choice of condenser is important and many mains suppressor types are not good at resisting voltage spikes in the low tension circuit (which can be surprisingly high). Although many rolled paper items are 'self-healing', ie if the insulation inside them is perforated, it self-seals, but they can't do it indefinitely. In many cases, their voltage rating is marginal, or inadequate. That's when a capacitor on the points is most likely to be a handy solution, and in the event of any future worries, easy to replace.
On my own machines, I've run our widgets for more than 6 years on three machines, and for less time/miles on 2 others. All the mags have modern windings - and no problems in a total of over 60K miles between them. Of around 10,000 widgets sold, the failure rate has been negligible (unless people aren't complaining and asking for their money back!) I think we have seen fewer than 10 returns, of which several were actually OK.
So confidence isn't misplaced, but there are no medium- or long-term get-rounds for faulty HT coils, and it's a lot of work to have to take a mag apart again because a short-cut didn't do the trick!
