I'd have to say connector oxidation is highly unlikely in this case because I inspected all the connectors, they're shiny new as the 2017 box was NOS, bypassed power and ground straight to the battery, and had them in and out several times with quality contact cleaner applied and all checked for physically intact crimps.
It repeated the same behaviour several times, and each time I pulled out my multimeter (don't leave home without it) and pulled the stator wires at the box to measure for 130 ohms which always passed, as well as a few hundred mV signal on kick and stuffed them back in again. If it were corrosion, I'd expect that would at least give me some spark on kick. But one time waiting half an hour gave my spark back and I was able to ride about a mile until it happened again. At that point, came back with fresh battery, no help.
Trucked it home. Next day, purred like a kitten first kick and then started running poorly. Took it up the street, high RPM was good, brought it back quickly and couldn't keep it idling in the driveway, died, back to no spark on kick. Took it out a third time a week later after double checking all the wiring again, disconnecting and reconnecting, tightening spade connectors etc., made it about 20 blocks in circles around my neighbourhood and had to push it home the last two blocks after it just stopped dead again. Ran out of steam at the end of the driveway, started first kick, rode it a few feet, and it died soon after, wouldn't restart. Later that day, spark again.
The repeatability gives evidence toward thermal failure, just like a coil, but an electrolytic inside the box could do something similar, or even a semiconductor. Fullminator on Britbike.com had it happen twice with two different Boyers.
Thanks for the Walridge tip; that's a good deal.