In a tank or regulator combustion is prevented by O2 cleaning to remove any potential fuel. This means removing all oil and rubber products that might burn in a high O2 content environment.
There is no such thing as "preventing" combustion where there are soft parts present.
Brass will not burn to 10,000 psi of pure O2. However, Viton O-rings definitely WILL burn, and so will nylon 6/6 seats (which are normally considered "oxygen compatable." If you get Christolube hot enough (290F or so) it will decompose into flourine compounds (which are NASTY - as in one breath and you can die from it stuff.)
The ENTIRE trick is to not create the hotspots and to remove all gross (e.g. loose hydrocarbons, loose bits of carbon steel, etc) contamination. However, that does not PREVENT O2 fires, it just makes them less likely.
Slamming valves open is a VERY bad practice, and sooner or later you will pay for it. Most of the time, as noted, you don't get the "regulator in flames" deal, you get a flash fire and a major leak (the fuel source is immediately consumed and the fire goes out when there is nothing left that will burn.) The major safety gain with O2-cleaning is that the fire will go OUT (the fuel is limited to the O-ring or seat that combusts); what is REAL dangerous is if it stays lit long enough and with enough fuel to get a hose or something similar going, or if there is enough fuel in the tank or a path back to it to sustain combustion - then you get either a very ugly fire or potentially even an explosion.
There is enough fuel in the average tank valve or regulator seat to blow the seat plug clear out of the assembly all on its own.
Don't slam valves on, O2-clean or not, and keep body parts away from the potential ejection sites when you DO turn on valves (e.g. the end of valve bonnets and regulaor HP seat caps.)
For all intents and purposes any FO2 over 50% may be treated as pure O2 - over 50% O2 gas behaves under pressure close enough to pure oxygen that you may as well treat it as if it is.
