It is possible to use a Mid Side EQ process where the sum and difference signals are treated separately.Not only that, but the degree of the dip depends on the balance of direct and reflected sound. More reflected sound, and it fills in(partially, a real center is still better). So I don't see how you can compensate for this with a fixed EQ in the source.
I do this myself to try and combat the darkening of the centre image that results from having very little early reflections. Purposely adding this into a master wouldn't be a good idea but an unconscious adjustment to spectral balance could be applied.