Some problems with this idea.
What happens if the flaps fail to deploy on the t/off run? You may not have enough runway to stop!! and no flap to get airborne.
What happens if you get assymetric flaps? You most certainly don't want to get airborne in that config.
How long do the flaps take to run to F5 or F15, exactly?
Is it the same for each A/C in the fleet? This will determine how much runway we use before we get to the rotation point.
What happens if there is a delay in running the flaps on a short runway due say the PNF not hearing the command to run flaps?
How much performance is going to be lost in the above scenario?
How do you determine the amount of safety margin when writing the performance manuals with so many variables?
That was just a few very simple examples of why we do not run flap on the t/off roll. I didn't want to get into perf "A", which is a nightmare for most of us "drivers airframes" let alone anyone else. I am not saying you wouldn't understand it, merely I don't want to keep typing all day.
If you did use the suggested method for departure, you would need runways at least 20,000'long to account for all the unknowns and variables. Since I have
never seen a runway this long, nobody would be going anywhere. I like to keep things simple, that way even I tend to get it right
