Please find attached the review cited. Section II.E1 contains discussion on many attempts to determine the InGaN bowing parameters. The overall range is between 1eV and 3.5 eV, and eventually the authors of the review recommend a value of 1.4 eV, that is slightly higher than the default value used in SiLENSe.

Unfortunately, it seems now most people actually doing LEDs and LDs think in terms of the relation between their own grown parameters and the resulting wavelength, and Indium content they provide in papers is just an estimation (most papers just say "InGaN MQW emitting at ??? nm"). I have seen no recent publications aimed in clarification of InGaN bandgap variation. The reason is that to do it one need

- Be able to grow good InGaN layers with no (low) In composition fluctuations for a considerable In content range
- Be able to measure accurately the In content in these layers
- Have a good model to get relation between the PL spectrum obtained and real bandgap of InGaN layers

Each of the above points is very non-trivial task.