The examples that you are giving are of the "failure condition is static, and the gradually degrading building condition approaches & then reaches it." This is, I believe, exactly correct for the towers, and also for the progressive failure of WTC7, up to the start of global collapse.
But I don't believe it is quite correct for the start of WTC7's global collapse. I believe that it was more like the classic regenerative buckling seen when one put 40 pounds of pressure on an empty Coke can with ones heel & then reaches down & taps the side, getting the side walls to bow inwards slightly & the can to crush ... at near free fall acceleration.
The can wall is stable until the taps to the side. It wasn't gradually approaching collapse.
In the case of WTC7, the inside of the building had collapsed from floors 5 to 15 across the width of the building, the stronger, highly redundant outer walls were still standing, producing a shelled out "building on 10 story stilts".
The "tap" on the side walls was outward, not inward, and was provided by the internal falling debris which built up in the lower floors and (what a shock, I tell you!) did not want to stack itself neatly at the end of its fall. The debris pile put a sudden outward load on the lower, outer walls.
This is corroborated by an eyewitness who said that he saw the lower portion of WTC7 bow outward. One can also see this 8-10 story, 3 point buckle in NIST's simulation.
The progressive failure (prior to global collapse) was as you described, but I believe that the event that initiated the global collapse was a distinct, discontinuous event, like the tap on a coke can, got the global collapse started.
A small detail. Irrelevant in the big picture.
But thermite nor bombs produce multi-story buckling.
BTW, the east wall of the building was so sturdy that the lower 10 stories or so remained standing across the entire east side's width.