Mutt,
My reference to - "the same margins to Accelerate-Stop as presently exist for the continued Takeoff case"
In applying the present legislation, the second segment climb utilises Nett performance data, whereas the accelerate-stop manoeuvre essentially uses Gross data. If the relatively simple case of a 1.6% Net gradient is required for a 2 engined aircraft, the aircraft must be capable of achieving a 2.4% Gross gradient, a Net Vs Gross performance delta of 0.8%, or put another way the aircraft only need achieve 2/3 of it's Gross performance capability to ensure obstacle clearance, i.e. 50% above minimum required. At higher gradients, it is well understood that, as the decrement is fixed, this actual percentage margin decreases, but remains substantial. Again, in the 3rd segment, Net level flight is assumed to commence at the horizontal point where the Gross gradient achieves acceleration altitude, i.e. at a lower altitude in the same ratio to Gross as in the 2nd segment case - a substantial margin.
Both of these (2nd and 3rd segments) examples illustrate significant margins above assumed (Net) performance.
For the Accelerate-Stop consideration, Gross performance is used throughout with considerably lower margins. The only significant margin is the certification requirement of keeping one means of retardation in reserve, typically Reverse Thrust. For most aircraft with which I'm familiar, this typically is ABOUT 15% of the ASDR, well short of the 50% example quoted for the 2 engined aircraft. Just about the only other margin is the pilot's ability to react in a lesser time than assumed for certification purposes, something which is rarely achieved.
Longer runways are prohibitively expensive, and, as you've said so eloquently, we would simply reduce thrust accordingly to take advantage of it. I would, both as an RTOW creator, and as a pilot.