I don't think it's even refuted. It's just redefining what 'decision' is and using that to somehow make the case for free will.
Previously they were saying that the start of the neural activity was when a decision was made, and now they are saying that the decision is only made when a particular threshold is reached, but either way the decision is a result of neural activity that began before the person was consciously aware of it. They're just redefining terms to reach the result they wanted.
The assumption that conscious awareness is necessary for an exertion of free will is also a mistake. The original conclusion was chock full of assumptions. No one debated the specifics of the observed data, what is debatable is the meaning of that data.
Previously they were saying that the start of the neural activity was when a decision was made, and now they are saying that the decision is only made when a particular threshold is reached, but either way the decision is a result of neural activity that began before the person was consciously aware of it. They're just redefining terms to reach the result they wanted.