Ouch... I can feel the pains of poor planner... And hunders, hundreds of hours spent making it work... How do you plan on integrating it with LA and S-Curve?

Shouldn't the RE motor motion just be X+E if the motors are mounted like the attached picture and E-X if the X and RE motor are coaxial and facing each other?
Sorry but I have no clue about the inner workings of Marlin kinematics computations... :(

Without corrections - yes, in theory it's no different from CoreXY. In reality? At high speeds? I suspect extrusion will suffer after 70-80mm/s even without LA and Scurve.
Though your idea doesn't make sense anyway. If you put it onto XY extruder Z bed - it'll just make one axis twice heavier, thus it's a massive downgrade. If you but it on XZ extruder Y bed - bed is still as heavy as motor, thus it's just useless.

Yep, but it doesn't make bed any lighter, so this is no better than pancake with gearbox or belt

Bed? The aim is to reduce the mass of a X carriage with a direct drive extruder assembly.

What's the point reducing mass of only one axis? Your speeds and quality only as good as heaviest carriage.

The point is to get at least one axis better.
The bed stays at a constant height while printing; the X carriage is lifting upwards while printing and as a result you end up with a mass of ~200grams being accelerated back and forth at say 2000mm/s^2 at an height of 300mm from the bed surface if you're printing a tall model. The lighter the moving parts, the better... or not?

Yep, to an extent. Remember that while your light X carriage go up your bed become heavier and heavier... IMO, it's better to build proper rigid and heavy CoreXY and don't waste time on half measures.