It isn't merely asserted. An analysis was done which allowed for NIST assumptions of shear studs breaking on the beams, no shear studs on the girder, and no expansion to the east by the 52 foot long beams.
In case you didn't notice, I explained earlier that the girder would get trapped behind the side plate even if it is only heated a slight amount. Its western corner only has a 5/32" clearance from the side plate and that gap gets closed with a temperature increase of just 70 F. The beams need to be at 600 C (1,112 F) to get anywhere near a plausible amount of expansion and the girder was right next to them and also connected to them. The girder could not possibly stay at room temperature and would thus be in the envelope of the side plate under all possible conditions where the beams would be pushing it to the west.
You have a serious conflict since you insist on supporting the NIST collapse initiation hypothesis in general, but don't seem to be able to articulate a scenario that would work for it.
No one has shown that collapse would not initiate in the NIST scenario given the addition of the side plates or any other structural element omitted by NIST. Hulsey and Arup specifically tested alternative scenarios and found, in those alternative scenarios, there would be side plate trapping (and I'm being generous by characterizing Hulsey's conclusions to date as actual findings, given that his approach to local connection failures seemingly makes little sense). (Arup also found that in several of those alternative scenarios, the girder still failed after the side plate trapping.) But no one--not you, not Hulsey, not Arup, not Weidlinger--has recreated NIST's comprehensive model of the first 16 floors of WTC7 and tested, as NIST did, how the observed fires--acting in an actual progression scenario--would damage those floors. Reducing the problem to merely being whether the girder moves against the side plate at a certain point in time under a specific set of conditions does nothing to undermine NIST's report unless you show that also happens under NIST's scenario, taking into account the fire progression and the damage in the surrounding area.