Can I timidly raise my hand in the back of the room here and ask why on Earth anyone would want to place an omni coincident with a pair of X/Y cardioids? Any cardioid has 50% pressure response (= omni). Thus a pair of coincident cardioids--regardless of the angle between their axes--gives you a stereo pickup in which a MINIMUM of 50% of the output of the two microphones is the same signal. And that minimum case occurs if you place the cardioids back-to-back; when you set a more reasonable angle between the cardioids, the cardioid patterns overlap so that there's even less difference between the right and left signals. The more this cross-blending occurs, the closer all apparent sound sources move toward the center in playback.
If, despite being aware of all this, you're still going to add an omni to an X/Y cardioid pair, I have two suggestions:
(1) Apply a low-pass filter (say, at 100 Hz or lower) to the omni's signal before you split it out to the left and right, so that you're only extending the low frequency range of the X/Y pair--not utterly trashing the (usually already marginal) midrange stereo separation.
(2) If deep low-frequency pickup is your main motivation, be sure to use a pressure transducer--which some of the microphones mentioned previously in this thread are not. No dual-diaphragm, switchable-pattern capsule is a pressure transducer in its omnidirectional setting; that type of microphone won't have the flat, extended low-frequency response that you're looking for.
--best regards