The wrong value on the Y-axis when using lethargy units

This is only an arbitrary scaling factor that does not change the shape of the curve, which the lethargy representation is intended for.
As for absolute values to handle, what matters is anyway dN/dE, with its well-defined units.