A. We don't know, but I doubt that it will make much difference. If it does, then your results are probably too weak to withstand serious scrutiny.
B. I don't understand the "really move to be alligned [sic]" question. As for using motion.1D in 3dDeconvolve, you are probably right -- it wouldn't help much. But it probably wouldn't hurt much either.
If you have an experiment with significant intra-TR motion so that you want to try 2dimreg first, then you probably have bad data.