Did some trouble shooting today and
found the root cause!!!
Started from ground zero to double check everything I've done so far.
Cleaned battery positive and ground side terminals, ground to frame, positive terminal to slow burn connectors, removed wire terminals from plastic connector at slow burn fuse, removed terminals from power supply white connector in engine bay, cleaned buss bar and buss bar wire. In other words disconnected every connector cleaned and inspected.
All four Glow Plugs can be removed without removing anything besides the buss bar. Verified GP ohm values and even powered them to make certain they all glow bright red in just a few seconds of power. They stay red hot for at least 8 seconds when energized for 10 seconds. They do become hotter with multiple on cycles.
Verified relays function as they should. Used another battery, lights and jumper wires to make certain they work under light load. Even bought a new 40 amp relay just to make certain. Each relayed was tested using meter and physical test mentioned previously. Stock Relays are 40 amp sealed 4 pin units from Panasonic.
Verified battery voltage at the buss bar when GP activated. What? Why doesn't it want to start? Run a 12 wire from battery to buss bar wire white wire connector and it starts amazingly fast.
So where is the problem?...
There's only one place left to investigate with vengeance. In the relay socket itself. Visually it looks OK and there are no clues it is defective though.
Used a spare male terminal and inserted it into the four connectors on the relay connector. Two were tight and two were very loose. The tight ones were on the control (cpu side circuit, small diameter wire) and the two loose ones were on the power side of relay with the two large diameter wires. This has potential to be the root cause. But now to remove the relay connector from the metal support structure. But how to do so? Used mirrors camera, but nothing clearly visible on how to remove. Elected to slide a 0.010" feeler gauge down between the plastic and metal hoping it would depress the holding tab. Wiggling side to side while pushing downward finally set the housing free. Now to take the housing apart and remove the wires to see what is going on.
The white structure needed to be removed from the black housing by gently prying on locking tabs while trying to remove. It took a bit of effort but it came off. Now to get the wires removed from the housing.
Need to use two small screwdrivers to release the plastic tabs holding the wire terminal in the housing. The terminal/connector is not a common design one would find locally. There is little to no interference when a male connector is inserted to simulate the relay terminal. Tried bending the metal detail that provided the spring force, but after the first insertion it would relax. Ended up using pieces of a paper clip to provide the necessary shim to provide additional clamp force. Not the greatest solution, but it works. Fixed both power side terminals using this method. Reassembled housing, inserted relay with significantly more insertion/extraction force. Now I'm certain there is good electrical continuity between the relay wires and relay itself. If I had more time, more area to work behind the dash and it wasn't so cold out, I'd make up my own relay connector.
The design of these wire terminals is not something that inspires confidence. I'd be hesitant to use them even if I could find a source for them.
Additional comments:
1. The two power side female terminals found inside the relay housing are much wider than the 0.25" wide terminals found on 40 amp relays. They can definitely accommodate the wider terminals typically found on 60 amp relays. I don't think these wider terminals would matter because they are the same thickness and this is the dimension that is critical to generate clamp force to yield good electrical continuity.
2. The relays initially seemed to assemble into the housing with some force, but this was due to the two control side terminals. ie: It masked the poor fit of the power side terminals.
3. I'm guess there was just enough metal to metal contact for the meter to read battery voltage at the buss bar, but not enough to drive the high current demand of the GPs.
4. Purchased a 40 amp relay from Advance Auto for about $9. Currently using have this installed. The housing is slightly larger but it still fits as the pin spacing is identical. Auto Craft 40 Amp HD Relay Part # ACA715A
Advance Auto Parts - Down for Maintenance
5. The stock Panasonic relay is nothing special although it is sealed. No diodes, resistors, etc... to suppress voltage spikes when the field is collapsed. (Went through the effort to decode the part number on Panasonic's web site and relay technical info.)
Although the temperature wasn't that low when I finished (28 F). It is amazing how fast it starts after just one GP cycle. It seems to start within one revolution of the motor.
If it's much colder tonight I'll try in the morning and see what happens. At this point in time I'm very confident the root cause has been found.
Hope someone finds this info beneficial.