I think the article is pretty misleading but the conclusion is clear: "There's plenty of evidence showing that marijuana use impairs key driving skills. If you get really stoned and then get behind the wheel, you're asking for trouble.".
The problem is that the legal thresholds for THC are so low that "stone drivers" are not actually stone, making the whole federal study flawed.
Here: "Several states have passed laws attempting to define "marijuana-impaired driving" similarly to drunk driving. Colorado, for instance, sets a blood THC threshold of 5 nanograms per milliliter. But that number tells us next to nothing about whether a person is impaired or fit to drive. "
If you look at the data, "any illegal drug" is also not correlated with more accidents.