...
It can't be that hard. But yes, it can be confusing, has happened to all of us... then is simple and obvious when the demon is finally found.
The engine only needs three things to run. Compression, fuel/air and ignition. I guess we could randomly brain-storm and maybe somebody will hit upon a useful idea.
It almost can't be compression. With one caveat I mentioned earlier.
If it has a knock sensor that has an extreme ******-ignition ability, then it could knock and ****** back so far as to kill power, but I can hardly believe it would ****** that much. Unless the computer is in the heat. Or any ignition module, even new does that in your case. EDIT: I see the word r e t a r d starred out above.
So if not the above, that leaves fuel and air. Since there should be a dog-house beside the driver, one could drive without the dog-house and have another observer look for problems on-the-go. Except it may run so cool when open, as to be fine if heat is a contributor. Not that it wouldn't bake the occupants as the whole house heated up. But the problems might be that the throttle body is not opening. Offhand, I can't imagine why not. Can the EGR do anything weird on these engines? It's part of the intake. There isn't a soft rubber intake hose ahead of the throttle/mass-air that can collapse when warm... is there?
Other things that can be done with the dog-house uncomfortably open is to more directly observe what the vacuum is doing when the problem occurs. By more direct I mean it is easier to connect a vacuum shop-gauge directly to a manifold vacuum source. Otherwise one may run a longer vacuum line to the cabin with the dog-house on and also observe a dash-like vacuum gauge.
A similar thing can be done with a fuel pressure gauge. Even an old mechanical oil gauge should be in the range of fuel pressure, or any mechanical water pressure gauge that has a brass bellows. One would want to be be certain that no high pressure (say 45#) fuel leak occurred. I think at least some fuel lines have a built-in fuel-line schrader valve for shop testing. Highway Patrol sometimes tap it (state shop-added valve) for motorists out of gas. It would take a bit of similar jury-rigging to get a non-specific pressure gauge connected. This extreme effort wouldn't make any sense at all until it was insured that the fuel line from the pump and tank was not somehow overheated. Then such a convoluted fuel pressure test becomes a last resort.
Lastly, I should explain how compression can change. You probably have iron heads, but if the steel valve inserts in aluminum heads, as I mentioned earlier in http://www.rv.net/forum/index.cfm/fuseaction/thread/tid/30288794/gotomsg/30288823.cfm#30288823
come loose, the intake manifold pressure drops to near zero (or ambient) with the throttle open. But even with the throttle wide open, the engine is running under vacuum (or at least some cylinders) when the inserts stick. Some cylinders are 'throttled' by the insert ring following the valve out. When a piston compresses a high vacuum, such as at idle or other port restriction, it forms very little compression. This is the reason the throttle should be jammed open during compression checks, with all the plugs out also of course.
For those that want to understand all modern fuel injection, I highly recommend this book: https://www.ebay.com/itm/384450433528 . Charles (edit: not Fred) Probst helped develop modern F.I. ground up and then taught other Ford engineers how to develop their program introduced in 1988.
Wes
