Notes. (A) Yost et al. (2003); (B) Aksulu et al. (2020), assumed ISM profile; their values have been adapted to be consistent with the usual assumption of 100% electrons being accelerated; (C) Chevalier & Li (2000), no errors provided; (D) Panaitescu & Kumar (2001); (E) Frail et al. (2003), no errors provided; (F) Panaitescu & Kumar (2002); (G) Panaitescu (2005); (H) Harrison et al. (2001); (I) Yost (2004), Yost et al. (2003); (J) Resmi et al. (2005); (K) Frail et al. (2006), no errors provided; (L) Cenko et al. (2010); (M) Cenko et al. (2011); (N) Varela et al. (2016) (O) Laskar et al. (2018b); (P) Laskar et al. (2018a); (Q) Laskar et al. (2018c); (R) Laskar et al. (2019). ⁽¹⁾ The Panaitescu & Kumar (2002) modelling results hinge on the interpretation of the sudden re-brightening as an observer being initially outside the jet, and the related reddening as ν_c-passage, compared to the more canonical interpretation in terms of a late shell-collision (Vlasis et al. 2011). Similarly, also none of the other afterglow modelling attempts cover the clean afterglow (e.g. Wijers & Galama 1999; Granot et al. 1999; Frail et al. 2000). Chevalier & Li (2000) find consistency with a wind environment when fitting just the radio data and the R-band flux normalisation, but fix p. ⁽²⁾ ISM and wind environment models explain the data equally well; Frail et al. (2003) prefer the ISM model due to the lower ϵ_e. ⁽³⁾ The values given are for the narrow jet of the two-component jet model; the wider jet explains the data at > 1.5 days post-burst, and carries substantially more energy. ⁽⁴⁾ The wind interpretation is consistent with McBreen et al. (2010), though they derive θ_J< 1.​​°1, and did not constrain the microphysical parameters. Lemoine et al. (2013) assume the > 100 MeV emission to be synchrotron, ignore spectral slopes, and use a two-zone model with decaying micro-turbulence to infer the fireball parameters, thus going substantially beyond the standard model. ⁽⁵⁾ McBreen et al. (2010) noted that modest host extinction of order 0.2 mag is consistent with the GROND data and reduces p to ≈2.4; otherwise consistent jet angle and energetics, but again no microphysical parameters. Lemoine et al. (2013) assume the > 100 MeV emission to be synchrotron, ignore spectral slopes, and use a two-zone model with decaying micro-turbulence to infer the fireball parameters, thus going substantially beyond the standard model. ⁽⁶⁾ Laskar et al. (2018a) prefer the ISM solution with the argument that the wind solution overpredicts the early radio data. However, the wind solution fits the X-rays much better, making it a similarly viable option in our view.