E. Prompt Polarization
In Section IIA4 I discussed the detection of very high linear polarization from GRB 021206 . While the data analysis is uncertain several papers claimed that this detection has strong implications. First Coburn and Boggs  suggest that this polarization indicates that the emission mechanism is synchrotron. Lyutikov et al.  and Granot  suggest further that it implies uniform magnetic fields within the emitting regions and the first even conclude that this implies that the relativistic flow is Poynting flux dominated and that the dissipation is in the form of external plasma instability. Waxman  and Nakar et al.  show however that: (i) Random magnetic field in shock's plane could produce almost as high polarization as a uniform field (provided that the emitting jet is narrow and one is looking along the edge of the jet). (ii) Even if the magnetic field is uniform the flow does not have to be Poynting flux dominated. They also stress that while in the uniform field case we expect high polarization in almost every burst in the random field one we can expect high polarization only in very few bursts. The different time dependence of the polarization  could also enable us to distinguish between the two possibilities.
Lazzati et al.  and Dar and De Rujula  suggest that this polarization implies IC (which can have in principle higher intrinsic polarization). This shows that even the simplest conclusion (that the polarization confirms synchrotron as the emission mechanism) is uncertain. My overall conclusion is that without further data on other bursts (which is, unfortunately, quite unlikely in the nearby future) not much can be learnt from this tentative detection.