I would say this is the correct annalysis:
"Or would it be concident with contamination of the crimped connection (oil wicking into the crimp join between the strands by way of capillary effect"
Especially if there are traces of H2SO4 or H2CO3 (sulfuric acid, or carbonic acid) in the oil. Condensation overnight plays a big part too.
Try taking two duplicate, brand new cables, remove the insulation from the middle of one of the brand new cables, leaving a good 6 to 12 inches of insulation on the ends, then dip the exposed middle wires in water (use a plastic bucket!! so as not to creat an additional ground), or oil and measure the conductivity across the wire and compare to the fully insulated wire. You could even run a load test that heats the wires up to compare loaded conductivity.
Be sure to test the two wires first (before removing insulation) to make sure the two insulated wires have the same conductivity, or do the load tests first on them. The load tests would be a more precise test.
Take say a pair of very closely matched 100 ohm high power resistors on one end of each wire, then power the wire up across a 12 volt battery. Even better, maybe use a 12 V head light bulb. Use the same bulb and battery to test both wires, or load test to find two matching bulbs.
All other things being equal, (equal battery voltage, equal light bulbs or resistors) I doubt you could measure any difference in the two wire sets conductivity or back voltage due to the oil or water contamination of the middle wire bundle.
But that same contamination if it wicks (capillary action) its way to the end connectors can cause corrosion at the wire to connector contact
points in just a few weeks especially if it is corrosive oil like battery acid fumes and that will cause loss of conductivity, or old overused oil.
For others reading this that don't know, resistance is inversely proportional to conductivity.
R = 1/C
C= 1/R
I would find it interesting someday to see if a wide flat thin strap carries more current than one with the same wire size and wire count as a round cable bundle, due to spreading out the surfaces of the wires. In other words, at some point the geometry of the cable affects the surface area relationship of the multiple wires to each other.