"Now how about including a non-operative spare or 2"
Could be done. But that would take more software to support, what I was describing is a small script(for the e-mail) and existing drivers/firmware. That could easily add another 3-4 months to the time to market. Probably worth it, though.
SATA drives are currently available at 400GB. Given that 1 hour of decent, compressed video is about 400MB, that means there is about 1k hours per drive. Say your RAID set is 4 drives, that means you have 1.2TB of storage, assuming RAID 5(capacity = size of smallest drive * total number of drives - 1). That gives you something like 120 days of video storage at 24 hours a day. With a standard 8 disk SATA RAID controller(with one drive for spare), this means you can have 6 thousand hours of stored, high quality video. Or about 4 thousand movies.
So what does this mean? At a guess, the home theater crowd could be serviced by early 2006 with a system for about $5k giving the full 6k hours of storage and at least 4 channels and probably 8 of 802.11g. The cost of development is hard to determine, but we are looking at at least 4 people for 6 months for a system that doesn't include the hot spare. The hot spare option is probably closer to 6 people for 9-12 months. Still, for an American company with good people, the cost of development should be not a lot more than $2 million, probably a lot less if the developers are willing to put in "sweat equity".
i.e. it seems like it is a no-brainer. But, given that no one seems to be thinking along those lines, it might be more complicated than that...