Oz, I think that's right, it couldn't be wringing force because unscrewing would put it in shear, where it is weakest.
Likewise I don't think it's any kind of suction force because of the same way it is being attacked -- suction can only be 14.7 lbs/sq in, which isn't much whe figuring the area involved, and it's oriented the wrong way to resist turning -- just like a suction cup on oiled glass -- you can slide it sideways relatively easily. And in addition, the threads apply mechanical advantage to breaking the seal in the axial direction -- plus the advantage of a 3 foot bar Mick used as a lever.
I don't think its suction or wringing.
That leaves galling in either register or threads as the only other possibilities. Since it went on easily, was oiled and cleaned, and wasn't tightened hard or worked while turning, the most likely problem is that something jammed in those areas that wasn't seen or felt, and did not inhibit screwing the chuck on, but only acted when trying to remove it.
For that mysterious circumstance, I can imagine a piece of thread with a crack in it, maybe in the backplate, not being noticed and staying in place when the chuck was screwed on. That would explain the ease. But then when unscrewing the thread piece is forced outwards and jams, like a latch lock. As more unscrewing pressure is applied the thread piece, being relatively soft, deforms in length, increases in width and becomes more like a plug. Expanding and locking the thread further with every increase in force.
That's about all I can think of considering the conditions things started with -- which would seem to have been ideal.