Not very important for practical purposes. The purpose of the test is to determine if a AI program is capable of fooling a human interlocutor to take it for another human. These days, useful AI systems tend to be narrow ones that fulfill a well-defined purpose applying machine learning to big data. Their scope will most likely broaden in the future. But for what purpose would the Turing test be a useful benchmark? AI systems are not created to fool people into thinking they're interacting with a human instead of an AI.