My head hurts thinking nowadays. It's 25 years since I was last tested on this stuff and never used it since.
Two 120degree phases connected will partially cancel each other out resulting in a more of a weak AC pulse than two individual sine waves. This would leave the rectifier with not much to work with. It could rectify it but the result would be lower voltage and a DC that would be almost impossible to smooth.
I think the biggest problem would be at the stator and plug side. I have not done short circuit testing on AC outputs but theoretically, there will be a current surge between the two windings which dies away. I assume this would happen upon each start up if there is a permanent short across the phases in the plug or each time oil builds up to form a short.
It is most likely this current that fries the plug and heats up the insulation on the phase wires. In addition to the usual heat that comes up the wires from the stator when it's hot and the rect that runs hot anyway.
Following this along, short all 3 phases and you should get an even larger current surge followed by just about no AC current arriving at the rectifier.
My theory does not explain why a rectifier may be damaged by the shorting in the plug. My plug melted but my rect was fine. I know some have said it killed the rectifier but someone else will have to come up with why. Maybe some of this brief current surge is felt by the rectifier diodes and its all a bit too much for them?
Like I say, my memory of that stuff is hazy now, plus in the day I did enough to pass and spent the rest of the time drinking beer like all students should. Especially as I did my apprenticeship and tech quals in the Royal Navy, beer drinking was a compulsory part of training.