Hi Christine,
In the example you put forth, I totally agree that a 3dANOVA4 would be practical and useful. Luiz's example is a good one too that would make good use of a 3dANOVA4. My only fear is that people will decide to complicate their designs further knowing there's a 3dANOVA4 or 5 available (i.e., putting what should really be 2 separate experiments into one design in order to save on limited scanning time). I've seen this sort of thing in the past, and helping them make sense of the results was a nightmare.