One way of looking at this is to ask what the mirror provides that a mirrorless solution doesn't? An optical viewfinder and phase based AF are the obvious answers. EVFs are getting better and if the technology behind the main sensor phase AF as seen on the Nikon 1 system can be implemented on APS-C and larger sensors then there would seem to be no compelling reason to retain the complication of a mirror system. If main sensor phase AF can't be implemented on large sensors then there's always the pellicle soluton. Mirrorless cameras certainly offer the potential
for small size but, as I argued in Size Matters
, for creative effects based on depth of field the diameter of the entrance pupil of the lens doesn't change whatever the size of the camera/sensor. Similarly with low light situations. So while the marketing folks may concentrate on smaller size as a way of selling these cameras I would suggest that for a number of situations beyond the usual "happy snapper" ones there is no significant size advantage at all because of the glass needed.
I think there are two main factors holding back CSCs (compact system cameras) at the moment. A limited selection of native lenses and, crucially, the high prices. It's not at all clear to me why a camera which is mechanically simpler should cost more. R&D costs should be no greater than for any new DSLR model and dispensing with fiddly mechanicals should reduce the parts cost and assembly cost. Maybe the manufacturers either think we are so stupid that we believe we should pay more for less or they are trying to avoid cannibalising their own DSLR sales too rapidly? Or am I missing something?