Some of the descriptions are correct, some are not. The spudwiki page has some info on how the disposable flash circuits work.
(http://www.spudfiles.com/spud_wiki/inde ... mera_flash
An excellent page describing several cheapo camera flash circuits is http://www.talkingelectronics.com/proje ... asher.html
The trigger cap is generally 0.022uF or 0.033uF (22~33 nF) low ESR. The energy in the cap is (1/2)CV<sup>2</sup>. If charged to 300V there is (ignoring losses in the trigger tansformer) ~1mJ of energy in the trigger cap. The minimum igntion energy for propane in air is roughly 0.5 mJ so the flash trigger setup is pretty marginal as an ignition system. Sometimes you can get it to work but you really need a bit more energy in the cap.
In some circuits the ready light is across the trigger cap but the cap is still charged to 300V, not 60V. The neon light does indeed limit the circuit a bit. Often though the ready light is wired across the main cap instead of the trigger cap. Even when it is wired across the trigger cap it is done via a pretty large resistor which drops the voltage before it gets to the neon bulb. So the neon bulb is seeing (dropping) 60 or so volts but the 300V source voltage is also being dropped across a resistor as well as the neon bulb.
Often, if you measure the voltage across the trigger cap with a DVM you'll get a voltage in the vicinity of 60V. That is not the actual voltage on the cap. The cap is up at 300V but it is being supplied via a very high resistance resistor, typically a couple hundred Kohm to a couple Megaohms. When you try to measure the trigger cap's voltage your DVM's internal resistance is similar to the limiting resistor so you get a voltage divider. The DVM reads much lower than it should because the DVM draws significant current compared to what the limiting resistor can supply. A typical DVM has input resistance of from 1 to 10 Mohm.
Anyway, back to the problem at hand. The typical trigger cap is a bit small to be used as a spudgun trigger, there is just barely enough energy in it, any losses and there isn't enough. Increasing the capacitance will help. If you could boost the cap by a factor of two that might
be enough. A factor of ten would be better. You do have to worry about overpowering the trigger transformer though. If the core of the trigger transformer is operating near saturation with the standard cap then boosting the capacitance won't help. (I have no idea how close that transformer is to saturation.)
The other approach is to use the stock cap and set the spark gap width correctly. Since the energy in the flash trigger is marginal to begin with you need the gap set just right. The output voltage of the trigger transformer is limited by the gap width. If the gap is too small it breaks down before the voltage gets up to the 2KV to ~5KV the circuit is designed to produce. When the gap breaks down prematurely you get a longer duration lower voltage disharge. You never get quite enough energy in the gap to ignite the fuel, the energy dissipates too fast and you don't get ignition. If you can get the gap just right so that it doesn't break down until the trigger reaches it's full output voltage (and power) then you get a short duration high energy spark that can ignite the fuel. The tricky part is that if you make the gap a tiny bit bigger it won't spark at all (and may well fry something on the flash circuit board).
You can see the same behavior sometime with BBQ piezo igniters. A very small gap between very pointy electrodes often won't ignite a fuel mixture. Increase the gap a bit and it ignites consistently. My long handled butane lighter "poppers
" are very sensitive to the gap width. Too big gives no spark and no ignition. Too small gives a spark but no ignition. It has to be just right to work consistently.