Remember also that:
1. Most runways are longer than required.
2. Most take-off's are done with reduced thrust which have margins built in.
Obviously a short haul pilot!
I agree that a lot of times a mindless rotate is the reason for tail scrapes but I disagree with the rationale that the solution in avoiding them is to use a poor technique simply because the runways are longer than required and there is fat built into the 'p' charts. I'd prefer for the right technique to be trained and flown so that when you are confronted with the worst case scenario (and goodness knows it's always the worst situation when it occurs) it all works as advertised. Flying the incorrect way in QF on the 767 or 737 will do the F/O or Captain no good when they transition to the A330, 744 or A380 and are often taking off at MTOW and max thrust and anything less than the right technique in the event of a mishap can mean the difference between flying away and augering in.
All the docs I've read indicate that the overwhelming majority of tail scrapes happen due to a fast rotate from the very start (poor technique) rather than from a 'snatch' through the dead spot at about 8-10 degrees with decreased elevator effectiveness. The tail scrape with SQ happened because they rotated 30 knots too slow. Using the SQ example is somewhat mischievous as it's a far more complex issue than simply the rotate method.
Further, I use the full headwind component to calculate my derate- as you say there is fat built into it- to go easy on the engines. I also tend to not over extend the downwind limit unless it's a 'flukey' wind or light and variable. Don't read into my comments that I'm in any way relying on the 'fat' to get my way out of trouble I've spoken at some length to performance engineers (who happen to be pilots as well) and so I'm comfortable with not having to add extra layers to what is an already multi layered defence.