I suggest a flaw in analysis which is the apparent segmentation of the Upper Air profile, in the Arctic, the real action is happening between 1000 and 650 mb, segmenting upper air levels 850,700, and 500 mb reduces resolution and misses some real
important air volumes at higher pressures, where anything can happen between 1000 and 850, 850 and 750 mb etc... Raobcore profile looks a little unrealistic, and it is likely that the models are more correct for the lower troposphere than Raobcore.