hi Jana
as to 1) yes, you should use the recommended polort. generally speaking, these are good things, in the sense that they account for variance in the signal that 1) is not likely to be related to your task and b) would otherwise end up in the error term, reducing your power to detect effects of interest.
for 2) I think Rick or one of the more expert people needs to deal with this question, though it is worth running the script with -polort 5 and seeing what happens. you might try:
1d_tool.py -infile Xmat.1D -show_cormat_warnings
to get a better idea of what the problem is.
James