A good point. On the other hand the Alpha 55 correspond to 36 MP when recalculated to full frame.
I have noticed some aliasing/moiré on my Alpha 55 so I guess that it may lack AA-filter.
As a side note I did a comparison of Michael's samples upscaled to same size (with Pentax 645D as reference), in my view the picking order was quite clear:
Sony A900 and Leica M9
Sony A55 and Sigma SD1
Now, those images were intended for display on the web, not really optimal for this kind of comparison. I also noted that the picking order may be affected by viewing distance.
In general I'd suggest that increasing resolution is a good thing. Aliasing is reduced, AA-filtering is always optimized for pixel pitch, so smaller pitch means weaker filtering. You are right that the AA-filter may be removed, but my sample from the Alpha 55 indicates that artifacts can arise. By the way, that image is taken with 400/5.6 and 1.4x extender at f/10, I did not really expect moiré but I think it's quite clearly there.
The Foevon sensor would of course not have any color moiré, but it could still have some jaggies and false resolution artifacts. An example of that is clearly visible in DP-Review test of the DP2, also enclosed. We can see that the line pattern turns mush around 1800 but shows fake resolution at 2000 and up, this is an example of contrast inversion. It is possible that the effect is not visible on normal pictures.
Well, the key is the + at the end of 30+. current rumor for Sony is 39mp. At that fine of pixel pitch, the AA filter may not be necessary, so even if it exceeds lens resolution, a big improvement.