I think it is certainly far too basic, but Hermann Gruber's answer shows that we might be able to save the question: just rephrase it and ask what is known about the complexity of the problem.
It might be also fun to know what are the open questions related to this problem (papers as recent as 2008 might suggest that not everything is known...). Or if we can find an interesting subset of regular languages that admits a much more efficient complement. Or why we are still using regular expressions to express regular languages, even though they are so cumbersome to complement...