I always calibrate my light meters to my DSLR because it's the easiest to judge (what with the histogram and all). I usually first calibrate them to an 18% gray card, and then pick a few other scenarios around the house to double check that they agree in both bright light, and dim light scenarios. That may or may not be "accurate" in a scientific sense. Rather than strive for extreme accuracy, I just go for consistency. So long as all of the light meters I use agree with each other, I don't have any problems switching between them. In most cases, I would think that a good digital camera should be pretty close. But I wouldn't be surprised to find one off by a third of a stop or so. That being said, what's important is you find a solution that produces good results consistently for you. And that relies as much on your process as it does on your equipment.