I like this paper, too bad it took such a long time to get it to its final stage. I was therefore going to recommend to publish as is to speed this up, but I feel a few corrections and possibly a minor revision are required so I go for publish subject to technical corrections.
To start with the need for minor revision: section 3.4 looks to me as quickly put together without proper editing. It is rather sloppily written and contains a couple of rather odd mistakes (see below). Maybe my personal bias but it also has one of the more interesting conclusions: something is missing in the NOx chemistry. The same conclusion is derived from different angles in the paper by Legrand et al, and the recently submitted manuscript by Savarino et al. I for one would emphasize this in the conclusion section but it is sofar only mentioned as almost an afterthought.
Specific points (numbering refers to the line number in the final manuscript (version 2):

194/325: talks about a 4m inlet, but this is never described or referred to (I presume it is more clearly indicated in Frey et al 2013). Please give more info.

451: I find it odd to call NO3- concentrations in snow an "ancillary measurement", to me it seems to be a key measurement for this paper. I would make it as a separate subchapter, even if the current authors did not make the actual measurements

454: "accurary is larger than usual". I guess it is correct in the sense that accuracy is used here. But at first it gave me the wrong impression that the accuracy was "better" than before. Maybe change it into "the results are less accurate than before".

671: section 3.4, as mentioned I recommend to edit the whole section. To me it gives a feeling of handwaiving with numbers.

683: "- the AMFs are incorrect" ?? How so?

694: "So, if... then..." My first comment would be "yes, but if not... then not..." To me there is very little indication whether the procedure makes sense. I do not suggest to come up with a better one, but formulate it more neutral like "based on the latter observation, we decided to divide the dome-C observations by a factor 3 to arrive at a first estimate of 2-3 pptv"

698: "XO = BrO, ClO" Now here is a jump?! If you have ClO data then you better show them. But I suspect this is just a remnant from another manuscript.

703-706: again, not wrong, but certainly a rambling sentence.

715: "9.6 (3.3)". What does the (3.3) refer to? Maybe you mean when 100% of HO2NO2 is interfering? The earlier text makes a case that 25% is the likely factor based on temperature and lifetime in the CLD cell, so why suddenly bringing in the possibility that it should be 100%? Be consistent please.

717: "(not measured)" ? no kidding. This is mentioned far too often, just refer to earlier text where this is discussed, including the likely factor of 25%..

1026-1030: conclusion on BrO: firstly saying that "the data suggest 2-3 pptv" is too strong for my taste (these things have a habit of being requoted in future papers and start a life of their own). I would be more comfortable with not quoting numbers but referring to the actual text that it is "probably low, certainly less than 5 pptv". Furthermore the text with referring to the Theys et al and Salawich et al papers fits more properly in chapter 3.4. And once again, this harping on the potential HO2NO2 interference.... I have not tried to determine how often the potential HO2NO2 interference is referred to, but it sure seems to be invoked frequently. I recommend to skim to manuscript for it and cull the text. After all you talk about a missing measurement that in the end appears not even to be so important anyway.