Mutt's comment - "Maybe this airport proves that the concept of just using runway ends isn't always the correct approach" is very valid, and J_T (as usual) beat me to the response.
In my performance work, I take a conservative approach to what I term "Composite Slope" runways. I stand by for incoming klak for being too conservative, but this is the technique that I employ in generating RTOWs -
For each runway, I run 3 analyses for each Takeoff consideration, i.e,
(1) The RTOW for the Mean Slope,
(2) The RTOW for the greatest UP slope, and
(3) The RTOW for the greatest DOWN slope.
The applicable limit (Accelerate-Go, Accelerate-Stop, MBE, etc.) is then the LEAST of the limits found. Many pilots don't notice, but one management pilot asked my why, for the same runway, the Slope printed on the RTOW was different for Full Length Vs Intersection departure - The answer; composite Runway Slope. Now throw in a few other factors such as CWY and SWY slope differing from the "Main" Runway slope (The EPWA case), and the runway becomes interesting.
My "mind set" began when once holding at the takeoff point waiting for an aircraft (B747) taking off in the opposite direction (Actually Mutt, it was one of yours, definately no criticism of you). It was (and is), a significantly humped runway, both ends the same elevation, thus Zero mean slope, but significant UP slope to the centre, followed by significant DOWN slope to the end. The said aircraft faced a degraded Takeoff performance, due to the actual Takeoff being UP slope instead of Zero, and a significant detriment to Accelerate-Stop due to DOWN slope if such a thing had happened (it didn't). The main gear cleared the runway end by just a few feet.
Yes Mutt, I do agree with you, the concept of just using runway ends isn't always the correct approach, there is an alternative, however conservative it may be.
Regards,
Old Smokey