I can think of several reasons - different sides of the press are not averse to being selective of their sources, depending on which conclusion they are pursuing, for one thing. There are strong industry interests at play here - the producers of insecticides want to find that they are not guilty, the big bee-keepers want to hear that it has nothing to do with the way they cart bees around etc. So, you cherry-pick your data.
Secondly, it is often seen, in long-running illnesses and epidemics, that there are periods of remission before it starts going the wrong way again. If bee-numbers are up this year, that may be all it is; we will know in the coming years.
I think the truth is that we are seeing a long, slow decline; we won't lose all honey bees in the world, but the industrial scale bee keeping, particularly in the US, will be severaly challenged, and will probably have to change their business model fundamentally, from carting their monocultures around with a heavy load of varroa mites, viruses etc, to being much more locally based. It has for many decades been a common practice to rely only on a very limited number of bee strains with specific properties, like high productivity, low swarming and low agression. It isn't really a surprise that we now find all bee colonies susceptible to emerging diseases, I think. And, of course, queen bees have been posted all over the globe, helping the spread of infections.
This is just a minor part of the more widespread problem, that originates with the industrialisation of agriculture: the tendency to have enormous estates of monocultures. The chemical industry are one of the major culprits in this, in that they have made it possible to mask problems with insect plagues and depletion of nutrition; we must, by necessity, come to a point where these things no longer are effective, and then it is likely to come crashing down. A sensible way out of this would, in my opinion, be to get away from gigantic monocultures and possibly also commercial production for global export.