Well, to be precise, a neutron bomb wasn't exactly the planned end result. It is essentially a failed atomic bomb, releasing many neutrons from the fisson of many nuclei, but not enough for a sustained chain reaction. But some sharp dude realized that though it wouldn't yield the megatons to level structures, the neutron flux would kill many living things nearby the detonation.
An antimatter bomb (photon torpedo) is a totally different matter. Here, I think that if one could collect enough antimatter (which at this point is not [yet] a possibility), and somehow contain it with electric or magnetic forces so that it won't touch normal matter unti you wanted it to, it would not have the possibility of failing like an atomic device, for I don't believe there's any critical mass threshold to be achieved before being able have annihilation to release the energy.
So? Why spend billions of dollars building a device that does the same job as a dirt cheap neutron bomb would already do? From Merchant's blurb up on top, it *seems* they want to build a device to deliver a massive amount of lethal radiation. But we aready *have* a device to do that, this being the neutron bomb. Perhaps it would be more efficient(probably) or perhaps smaller(probably not), but both storage problems for a live antimatter device and the cost to produce antimatter make me leery compared to the realitively simple process of making an ordinary neutron bomb, if that can be called ordinary, anyway.
Granted, Europe wasn't terribly pleased at the idea when the US wanted to deploy them way back when, but that political climate is all but disintegrated in today's current situation, in addition to the fact that I don't see how using antimatter in place of atomics removes any of the problems for deploying such a device, except for the additional cost to store yet another esoteric weapon.
I do not see a clear advantage for an antimatter-based radiation device. It might have the reliablity you suggest, but the flipside of that is you can't disable an antimatter device the way you can with an atomic device. Storage would be more hazardous than pretty much anything else in our arsenals, and as such could end up posing more a threat to the people building them than the people they want to use them on. Yield is clearly an advantage, but since radiation devices are, from my reading, generally considered to be used as tactical, not strategic, weapons, why would you need really high yields?