are GRBs beamed?

It was all GRBs today, with Maryam Modjaz (NYU) at coffee talking about this paper that finds a gamma-ray burst by its optical emission (that is, not by having a gamma-ray or x-ray trigger). Then at lunch, more GRBs with Modjaz talking about the GRB progenitors, about which we only have indirect information from looking at environments (and especially abundances). I asked my usual questions about beaming: We know from models of the emission that GRBs should be highly beamed, and that the optical should be much less beamed than the gamma-rays. But that means that there must be many GRB-like optical transients that don't have gamma-ray counterparts. None have been found! What gives? In general, the empirical evidence for beaming in astrophysical sources has always been weak (think of all the arguments about quasar unification), with pulsars an exception. Of course the detailed answers to this question involve the relative brightnesses, time scales, and spectral energy distributions of the way-off-axis emisssion, so there is quite a bit of theory to do to interpret the observations, but right now it is looking hard to reconcile all the theory and observations around GRB beaming.