Except for energies 
6 keV, calculations show that EPIC observations will be dominated by the diffuse X-ray background from the
sky. This can be modeled as two thermal plasma components within the
Galaxy (at kT = 0.05 and 0.12 keV; the former being unabsorbed, the latter
with an average absorbing column density of
cm-2),
and an extragalactic power law component (with a photon spectral index of
1.46 and the same absorbing column density as above), see Chen, Fabian
and Gendreau (1997; MNRAS). The former is known from ROSAT sky maps to
vary in intensity by a factor of two with location. It is difficult to
assess what fraction of the truly diffuse background will remain
unresolved at the angular resolution level of XMM, and hence what the
eventual point source detection sensitivity of EPIC will be.
An additional caveat to consider is the detection cell size, because the signal-to-noise ratio will depend dramatically on the extraction radius assumed. For the associated plots herein, we have adopted a region confined to the order of a PSF core.