The problem with most of the EVILs out there is that whilst small and even pocketable with a pancake lens, the quickly become ungainly when a kit lens is attached.
For EVIL to really take off their has to be a step change in the quality of image sensors, so that you can either get APC-Like quality from a SUB APC imager (and probably something smaller than Micro 4/3) or alternatively a step change in imager sensitivity and noise processing so that smaller lenses (with greatly reduced maximum F Stops) are viable.
I have a feeling that until there is this dramatic revolutionary (rather than evolutionary) improvement in Image Sensors EVIL will remain a much smaller market than pundits and product managers would like!
The EVIL sysytem is not intended to be a cheaper DSLR system. It is just a "smaller" system with picture quality comparable to DSLR. Therefore if they can make the small sensor (smaller than 4/3) with image as good as the APS-C, I am sure that they will also apply the same technolgy on the APS-C sensor the make the APS-C sensor even better. So If we want the EVIL to be as good as the APS-C DSLR, The EVIL sensor must be the same size as APS_C.
As for the lens, It is harder to design a lens to have total line pair resolution of APS-C for a small sensor. The lens will need 2 to 3 times more resolutin per mm.
We should try to look at EVIL and DSLR just like the range finder (with interchangeable lens , like Leica and Canon) and DSLR (like Exakta, Canon, Pantacon, Pentax... etc) in the old film days. It is two seperate system to fit the different need of user. Either system may be more expensive than the other. So EVIL should not be in the cost saving path. It should be on the quality path.
My point is ergonomic rather than economic although ultimately economics come into it. Most EVILs with any zoom are not as ergonomically ideal as DSLRs. The major selling point from the manufacturers pint of view ids that they are more "approachable" by the average user looking for greater quality, thus (the thought goes) we will create demand for another product. My observation is that in reality they are less approachable as they are less ergonomic when configured as most would like to use them, and in fact are no cheaper in reality.
With a fixed focal length lens they become very akin to a pocketable Rangefinder.
Going off at a tangent for a moment, we also need to think about how retake pictures. Since for an EVIL to be really compact the viewfinder paradigm needs to be ditched. Looking at the latest crop (Olympus PEN 3 family. Panasonic 3 etc) they are getting there by using touch screens but the interfaces are still too slow and the displays not of sufficiently high resolution. When we can get a high resolution LCD at around 300ppi (Apple "Retina" style) and a generations faster processor then all sorts of interface improvements are possible and I am sure will come. (This technology is probably a year or so away looking at developments in Sartphones and Tablets)
Imaging a fast touch display that let you zoom with a multi-touch movement (a la Apple iPad/iPhone) and then let you pick hyper focal point or let you choose the depth of field by touching a beginning and ending position or conversely let you swipe the area you wanted out of focus etc. This type of interface makes the EVIL a compelling tool. It is not there yet but it is coming.
One thing is sure the speed of development of fast low power processors )Moores "Law") is greater than developments in either optical Imaging or optical design. So (and some will cringe) we will see the ability for the camera to compensate for the inherent limitations in Optical Design and Imager Technology in real time (today we like to pull a raw image and make the adjustments ourselves, but this is a counter intuitive workflow, especially if one wants to capture the essential moment in time a la Cartier Bresson).
If one looks closer at current Imager Technology and the Bayer Mask it becomes apparent that this is a highly developed "Kludge". Something more like the Foveon idea makes more sense as it will produce less noise per color channel. I think that the only reason that it has not taken off is that more engineers are focussed of creating workarounds for the limitations of the Bayer Mask.
This harks back to the old school fight between Leica and Zeiss, of contrast vs. resolution. For the mass market the Zeiss school of thought won because with the technology of the time higher resolution (compensating for loss due to dispersion) won out over higher contrast (compensating for loss due to diffraction).
Ultimately change is inevitable just as the bulk of the Canon AE-1P + 35-135 Zoom customers of 25 years ago is now likely to by a $200-#250 Digicam. It is the mid market that drives oct technology improvements - economies of scale.
Art is in the Brain of the Beholder!