I hope someone who is familiar with the Clark wiring of that era can offer up some info on whether or not a ballast resistor circuit (or a resistor wire) was configured into the ignition coil circuit to help with cranking/starting problems.
Here's why I say that:
For a long time in things automotive, ignition systems were fitted with a ballast resistor in the coil primary circuit. The ignition coil was built to be rated at a voltage level roughly 1/2 of the system voltage of the vehicle. During normal running operation, the ballast resistor did the task of reducing the voltage fed to the coil to a level the coil could tolerate.
But, there was a second wire that led to the same + terminal of the coil in addition to the wire from the ballast resistor. The second circuit was fed from a circuit that would only be energized when the starter was energized.
All of this was done to overcome the phenomenon of losing spark at the coil due to the tremendous current draw of the starter causing a corresponding drop in battery voltage.
The second circuit was usually fed from an ignition key switch having an independent output terminal that was energized at the same time as the starter control circuit, but NOT electrically connected to the starter control circuit.
General Motors accomplished the second circuit by feeding it from a dedicated post on the starter solenoid.
FoMoCo did it with a terminal on the starter relay.
Ballast resistors fell out of fashion and the job of lowering the running voltage to the coil was done by special resistor wires, instead of the ceramic body ballast using a iron wire resistor.
We had some Komatsu FG20-11 trucks back in the mid 90s that began to exhibit the same symptom you have regarding that they would not achieve an actual start up while cranking, only when you let go the key would they begin to run autonomously.
When I began chasing this problem I discovered that the second circuit (current direct to coil + during cranking) was just hanging loose in space at the ignition switch, and there was no empty terminal on the ignition switch to connect it to.
Queries to KFI revealed that the brass at the company I work for had specified the trucks to be built with a aftermarket ignition switch (that would match what we had on our shelves) instead of the Komatsu standard ignition switch for that model Nissan H-20 engine. The aftermarket switch did not feature a terminal to bypass the ballast resistor during cranking and so after some years of use, the ignition circuit in the truck would gradually build up resistance, and the current draw from an aging starter would increase until a point was reached where voltage during cranking would fall below the level at which the ignition coil could make spark.
The fault was not with Komatsu, it was because some bean counters at my company decided they knew more than KFI and Nissan engineers and substituted an incorrect switch for the correct one.
I was able to secure permission from our "powers that be" (the same ones responsible for the switch fiasco) to add a Bosch mini relay to the wiring near the ignition switch to feed the second circuit (ballast bypass) to the ignition coil when the starter control circuit was energized.
It is not as simple as just connecting the starter control circuit to the coil second circuit because if you did.....the starter would still be engaged after the engine starts due to being back fed from the coil primary circuit.
Sorry for the length of post but I thought maybe some background info about ballast resistor systems (or resistor wires) could possibly help in finding a cure.