Good development of the discussion in my view.
[First, for those who may not be aware, Mutt is (was ?) an ops engineering specialist (in his previous life and looked after this sort of stuff for a fairly large operation).]
Second, the problem which is being missed a little is that there are two supposedly complementary (but, often, in opposition) things in play -
(a) Design Standards - eg FAR 25
(b) Operational Standards - eg FAR 121, 135
The Design Standard gives you an aeroplane which is certificated to be compliant with a bunch of requirements and capabilities specified in the Standard. In general, however, it doesn't address much about what, when, where and how the operator might or should go about doing operational things.
The Operational Standard addresses what the Operator is required to do as a minimum.
For performance, one generally uses the data output from (a) via the AFM to figure out how to meet the requirements of (b). It helps if (a) and (b) are compatible and, while this usually is the case, one often sees the odd case of the left and right hands not knowing what the other is doing.
your either planning/recognizing for single engine performance to exceed any all obstacle and navigational aspects of your flight...or your not.
Considering OEI one should certainly be doing. The relevant question is how do you go about doing that ? FAR 25 gives you some data but not much in the way of tools. FAR 121/135 spells out some requirements but doesn't give you much in the way of how to meet them. Enter ops engineering (or those pilots who have made it their task to get up to speed on the ways of doing performance work).
It's immaterial whether you can figure it out, or whether it's required, or whether your told to do it.
I'm not sure that I understand where you are going with this comment. Figuring it out is mechanistic, requirements are in the Regulations or follow from sensible governance considerations and "told" is a combination of airmanship and corporate/regulatory governance.
Simply put, if you can't meet Part 25 specs to fly around on one engine at any phase of flight, then you shouldn't be there.
Not quite the case but near enough for government business. However, FAR 25 only gives you some data - it doesn't bridge the gap between data and not bending the aeroplane.
do not specifically require an obstacle clearance analysis ... it is appropriate to provide information to the flightcrews on the safest way to perform such a maneuver should it be required
Now I'm only a dumb PE with a bit of airline flying experience - just how does one provide information on the safest way to perform such a manoeuvre without doing some analysis ? Sounds like black magic mumbo jumbo to this little black duck.
To accomplish this, the operator may develop the methods and criteria for the analysis of one-engine-inoperative procedures which best reflect that operator’s operational procedures.
I suspect that that goobledegook means "do some analyses" ?
Generally, published missed approach procedures provide adequate terrain clearance
they left out the bit which says "for minimum terrain requirements and AEO operations"
Published missed approach has a climb gradient requirement
this might require a reduction in weight but still only is considering AEO
Departure procedure for the runway has a published minimum climb gradient
if nothing else this alerts everyone to the idea that we have some serious terrain here ..
A special one-engine-inoperative takeoff procedure is required
often this will have little to do with the missed approach ?
There are runways that are used for landing but not for takeoff
the "why ?" is important but the inference still suggests the need for a closer looksee at things.
Gents. This tells me that OEI climb gradients are considered
ah, no, not at all. Main shoot down is how do you consider OEI for twins, three holers, and quads simultaneously without specifying a lot of information ?
Also one can refer to FCOM 3.05.35 Approach Climb Gradient (%) to determine any odd %, other than the standard 2.5%.
How do you intend to account for the progressive clean up from the landing configuration to the missed approach configuration to the enroute configuration .. with varying gradient capabilities throughout ? Not to mention the hills here and there and the possible need for turns ?