There are a few motorized stitching solutions out there since a few years. Most have been focused at gigapixel spherical view kind of stuff, rather than stitching a normal architecture shot. But I guess it's more about marketing and software packaging than hardware, I think someone could make a stitch-based landscape/architecture solution based on already existing products. It would not be that portable though.
When I did stitching I used a manual head, easier to carry and you no worries about running out of battery. Personally I found stitching to be a too mechanical shooting process to be really enjoyable (again very subjective), but I do recognize it as a very good and cost effective method to achieve high resolution images or wide angle views. You can't beat the quality of spherical stitching, using only the center portion of a high resolution lens. Many use stitching on tech cams (moving the back and stitch within the image circle) which has some advantages compared to spherical stitching (you can usually cover the same view with fewer shots), but I have actually never done it other than for testing, I enjoy the one-shot process much more even if one have to compromise with quality at times. Had I been shooting more indoor architecture I think I would have used stitching more, because you tend to run into situations when you need extreme wide angle there for practical reasons (can't back out through the wall). The attached picture is such a stitch I made a few years ago with a Canon 7D (had no MF system at the time), I simply did not have a lens with wide enough wide angle to document the whole room in a single shot (not a super shoot with bad light etc, but worked well as documentation which was the purpose that day).
I see post-processing as a necessary evil, but I don't particularly like it. The less time I can spend with it the happier I am. Even if there are effective stitching software I'm convinced the one shot images does reduce the time you sit at the computer. This has also been a major sales point for MFD in portrait applications, that the default result in terms of color and look is closer to the end result than a typical 135 system will deliver. As far as I understand despite the advances of 135 it's still the general consensus that this is the case.