I've done some experiments with the basic 'Brown' circuit on a home-built outboard motor using vintage J.A.P. Model '0' two-stroke with a Wipac magneto.
All testing was done with the motor in water to provide a realistic load.
I used values as follows: R1=1K0 1Watt, R2=2K2, R3=1K0, Q2=BC337, however I chose a TIP162 for Q1 which avoids the need for the shunt suppressor.
I wired the module into the ignition system with a two way switch so I could switch between contacts and condensor, or electronic.
Initially I used C1=0.47uF, and the module worked reasonably well, but was not significantly different from contacts and condenser. Notably, starting was more difficult and when running at slow speed or idle the engine would stop unexpectedly. It was always reliable at low speed when on contacts.
Changing R2 to 4K4 and R3 to 2K2 reduced the advance by making C1 charge more slowly, and it made starting easier, but high speed running was no better than contacts.
Changing C1 to 0.1uF advanced the ignition and gave significantly better high speed running than contacts. But C1 still needed to be a higher value to reduce the advance at idle and to aid starting.
I reverted R2 and R3 to 2K2 and 1K0 and set about testing different values for C1 from 1uF to 0.047uF.
I couldn't find a single value that worked sufficiently well at both extremes. It was one or the other. I considered adding a switch to add extra capacitance for starting but decided against it because the actual voltage on C1 is indeterminate and at some points in the cycle is negative. Having another capacitor with unknown charge suddenly appear in parallel could cause undesirable effects.
For experiment I added a 10K Lin pot in series with a 0.47uF cap and put them across C1 (0.1uF), thus providing some variable extra capacitance to manually adjust the timing.
The results were spectacular. Really easy starting and smooth idle with VR1 at 0K, very smooth running through the range by increasing VR1 while opening the throttle, and much better high speed running with VR1 at 10K when the additional capacitor has next to no effect.
It's all manually operated at the moment, but I plan to rig this to the throttle cable via a adjustable linkage to provide an automatic advance/retard for optimum timing at all speeds.