The histogram does not report the underlying image, it reports the results from your editing,
That whole sky area looks light grey to me (as edited); and not clean white. This can happen when trying to correct for an extreme camera over-exposure, with an extreme highlight-darkening adjustment. And that outcome is what the histogram is - accurately - showing, so far as I can tell.
It is often possible with negative film capture, to creatively make a positive virtue of overexposure, with an attractive "high key" treatment which is subjectively rather like what our eyes see when looking at bright lighting. With film, overexposure is a matter of degree.
But digital images show very abrupt transitions of tone and of hue, at the edges of their fully overexposed areas, which cannot be adjusted away. With digital images, overexposure of a particular part of the photo is not a matter of degree - it is a matter of yes / no.
I found the "bug" in my process... I was convinced that no edit was performed on that raw file, while I already imported an .xmp file generated by LRTimelapse that was applying a mask to all picture area with -2.5 exposure. So this is why the blown area is grey instead white and the histogram behaves accordingly.