Unfortunately I am not really aware of any literature regarding this, however a generalization of the Arkani-Hamed et al paper to the case of d+1 dimensions should be rather straightforward.
Let us begin by noting a few basic facts about the group theory involved:
The little group for a massless representation (this is the analog of helicity) of the Poincaré algebra is given by $SO(d-2)$, while the little group for a massive representation is given by $SO(d-1)$. The number of degrees of freedom corresponding to massive spin-2 is given by the symmetric traceless tensor of the little group, which has $\frac{d(d+1)}{2} -1$ degrees of freedom. In the massless case a similar argument leads to $\frac{(d-1)d}{2}-1$.
Now the point of the analysis by Arkani-Hamed et al is essentially to understand the theory in the UV, i.e. at energy scales much larger than the mass. To do this they try to decompose the massive representation in terms of massless ones, a straightforward counting exercise shows that in this case a massive spin-2 decomposes into a scalar, a helicity-1 vector and a helicity-2, exactly as in the 3+1d case. Using this knowledge it should be very easy to generalize the previous results, it is mostly a matter of carefully keeping track of the $d$'s. The vDVZ discontinuity, will still be there although the relative factor in the radiation-radiation and matter-matter interaction will depend on d, this can easily be seen by decomposing the tensor structure of the massive spin-2 propagator in terms of three massless one's corresponding to the helicities, a nice derivation for the $d=3$ case can be found in Zee's QFT book.
I hope that helps a bit...
This post has been migrated from (A51.SE)