|The authors have submitted an improved version of their article and I favour the publication of the present manuscript.|
However, I would suggest a few improvements before final publication.
1. I would suggest to add “uniform” in line 7 of the abstract, i.e. “the effects of electric fields” → “the effects of uniform electric fields”
2. It is true that the investigation of electric fields higher than approx. 3 MV/m need further preparation. I agree that this would be out of the scope of the present manuscript; however, this could be something for further research (why not even include other groups?)
3. I appreciate that the authors added an overview of the processes involved (Table 13 in the Supplementary Material). However, this is not what I had in mind. I would have suggested figures showing the cross sections for the different models. Thus, the reader would have an immediate opportunity to compare his/her cross sections with the ones used in the models described in the paper. I understand that this might increase the size of the Supplementary Material, but it might be really useful. If the authors consider this to be too much, I would suggest to at least present the equations (and all needed parameters) where analytical cross sections are used. In any case, I suggest to reformulate lines 4-5 on page 9: “However, we will not describe comprehensively all the processes, models or cross-sections used by the different codes, and a table summarizing it is provided in the supplementary material document, section 13 (including all references).” => “However, we do not describe comprehensively all the processes, models or cross-sections used by the different codes, but provide a table summarizing all implemented collision processes in the supplementary material document, section 13 (including all references).”
4. I agree with the authors that going from the (average) friction force to stochastic codes is not straight forward. Maybe, it would be good to include this explanation into their discussion of Fig. 1 (maybe even of Fig. 5.3. of the Supplementary Materia). Hence, the reader might not get confused with this very low probability. In their reply, the authors say that authors researchers should try to reproduce the 12% run-away ratio. Why do they not write this into the main manuscript?
5. I have one more question about the runtime. In the previous reply, the authors write that the simulations run until i) all electrons are absorbed (only the seed electrons???) or ii) there are 20 (or 10,5,1) electrons above 1 MeV. Could it not happen that some simulations run forever? What if there are, say, five electrons above 1 MeV running away (1 MeV in a field of 800 kV/m is well above the friction force). They might never get absorbed, but not necessarily produce new electrons above 1 MeV. Why can this option be excluded? I am not saying that this scenario might actually happen, but I cannot exclude it, either. I would be very grateful if the authors could briefly elaborate on this.
6. In the reply to referee number one, the authors say “that we do not intend to validate/verify any code in this article.” This is a very interesting statement. If the authors only intend to compare the codes/models described in the paper and give some suggestions for comparisons, this is, of course, fine. However, I strongly suggest that the authors make clear that this article is not about validation/verification. That could be done in the abstract or in section 6.