Ok but given a presumably abundant quantity of combustible fuel that feeds the "excess" -- that is, the flare -- what was this quantity doing in the previous second? Why did it not ignite exactly when it was at the quantity, or the density, that feeds the later even flames? Why is it lying there waiting to feed a flare?
Would you say that the heat that is released when the fire ignites, reaches out to acquire more oxygen than subsequent flames have available? Or is something else going on?
It's curious that the change of state from non-fire to fire is explosive rather than gradual -- and yet it's gradual just before it actually combusts.
Remove the accelarant - and you will have a dramatically slow combustion. If you were a boy scout, and had to spend time rubbing sticks, or using a flint with kindling - you would see that under 'normal' conditions, fire starts quite slowly, and builds. What you are seeing is the application of accelerants - which "accelerate" the combustion. Again, light a candle, light some leaves, light paper, cardboard or kindling. Without an accelerant, it's a slow, blistering and painful process.