This is a small study on one apiary with a low number of hives. That's not evil or anything ... the results are what they are but they're not sufficient to reach any conclusions sufficient to shut down an industry.
What they do justify is application of the approach, with more rigor, multiple study sites, and a lot more colonies, to see if the results hold. On a small scale like this it could also be that the hives with high mortality had something else in common.
I've seen some of Bayer's data. I think they're pretty good at knowing immediate toxicity levels. They may have developed some skill at asking questions they know the answer to and like the answer. They might avoid questions where they don't like the answer. This could turn out to be one of those, but the results must be replicated on a larger scale. We've had a lot of small-scale studies that thought they had found the answer to CCD that wouldn't replicate.
This doesn't change my personal approach to pesticides. Our bees are far away from agricultural fields, and we're glad of it. The only pesticides we use are Bti mosquito dunks in our rain barrels and a touch of foaming insecticide down into carpenter bee holes.