You are correct in so many ways, about the load-shifting, how it tries to transfer load-stress horizontally across a floor, so that that load-shifting and continuing vertical support below undamaged areas of the same floor, will keep that floor “upright”, and how the general result of that dynamic will tend to focus the total affect of the load-shift to the core. Yes.
But, it was the combination of the nature of some of the catastrophic damage that took place, and the unique nature of the structure of the Twin Towers that MIGHT, taken together, have both contributed to the initial-collapsing floors and also worked to concentrate the weight of the collapsing load towards the core.
The uniqueness of the Twin Towers was that unlike most (most all, I believe) similar looking high-rise buildings, the vertical support design of the structure WAS different.
With most all other high-rise buildings, the outer-most vertical elements, that are clad with an exterior finishing material (the curtain wall you see when you look at the building) are but one series of many such series of vertical supports, from the outer-most wall to the inner core of the building.
That was not the case with the Twin Towers.
One of the selling features that the planners pitched with the Twin Towers was that the tenants’ space (and laying it out) would NOT be encumbered by dozens of vertical support pillars running through the space between the windows and the building core.
The outer-most wall (the steel girders it was made of) WAS the “other” vertical support besides the inner core. There were not multiple sets of vertical pillars between the outer-most wall and the core.
However, and in fact, due to the “load-shifting” calculations resulting from that design, a contract for specially designed and specially hardened steel girders was given to a Japanese company, just for the Twin Towers, because, until the Twin Towers, girders to those specifications had not ever been made.
The other feature they employed, to strengthen the floors, help carry the load across a larger open (vertical support free) space, and assist in the horizontal load shifting, was that the steel girders of the “floors” were not a simple one-dimensional horizontal placement of girders.
They were constructed as trusses, where a “single span” consists of an “upper” and a “lower” member joined in multiple places from end to end by diagonal uprights between the top and bottom of the truss. Just think of two parallel lines with a sort of continual “z” configuration running between them. That’s a truss. And the trusses are what joined the outer-most walls with the inner core (made the “floors”), with no vertical supports in-between.
I don’t think any engineers have confirmed that the design elements and how the structural elements worked and how they were affected by the particular damage they received was the cause of the very start of the “pancake” affect. But, as far as I know, no architects, since 9/11 have proposed buildings that employ the unique vertical design feature of the Twin Towers, for buildings so large - a single outer-edge vertical support system with no other vertical support until the core.
By the time the combined wreackage hit an undamaged floor it was heavy enough to break an unweakened floor, so the collapse continued down the building (You can see windows blowing out sequentially as it happens.)
Now you have the intact building section above the impact zone sitting on top of a hollow tube. Piston. Crush. The central building core guides it down.