Take it back to basics and imagine we're talking about a basic toy glider.
It works for that; it has to as they have no pilots! As does a large wing dihedral angle.
But both are fairly crude methods of achieving stability; no-one is saying that positive longitudinal dihedral ('decalage') is essential, just that it is one way of ensuring a measure of basic longitudinal stability in some types of aeroplane.
The movements of Centre of Pressure with the extremely small perturbations of which we speak would be small enough to be ignored.
'Downwash gradient'? Did they call it something else at CFS?
"Hey, Wilbur, why don't we just stick to bikes?"