** We thank the referee again for the review.  We really appreciate the quick turnaround!

Table 2: This added table is very useful. From the table, it is clear that some fluxes, that are listed in the main flux table/catalogue as limits, actually have measurements >1 sigma. I take it from the authors comments (in the response to the referee report) that this is because they have introduced a type of cascading prior on the main flux catalogue? I.e. if something is clearly not detected in a higher resolution (bluer) filter then it is assumed to have no flux in the redder bands, and the authors therefore force the redder bands to have limits regardless of aperture flux? Is this the case? If so, I think it should be stated in the text (indeed, had the table not been provided, this would have been impossible to tell). 

** Yes, that is essentially correct.  For a given telescope, based on our visual scrutiny we do not believe there are detections redward of a non-detection.  We have added this point to Section 3.6 where we discuss upper limits.

Additionally, the caption of table 2 still references upper-limits, which naturally is not the case. 

** Fixed.  Thanks.

Figure 4 & Section 3.4: 
The addition of the Planck 850um fluxes is a good one. The authors correctly flag that these fluxes may be systematically bright (and imply that truth is likely somewhere on the interval [SCUBA850,Planck850]) but I note that with the addition of these Planck fluxes the observed excess becomes much more convincing. Indeed, one can follow the excess by-eye from 500um to 850um in many cases now. I think the authors would be justified in drawing attention to this, as I think it bolsters their science claims. That is, however, only the opinion of the referee and need not be implemented. 

** We prefer to not explicitly highlight the 850um fluxes in the excess discussion, since those fluxes have additional limitations as described in the manuscript.

Equation 5: I briefly draw attention to a comment in my previous report, specifically that I thought the upper limits should be defined in terms of total uncertainty (cal & sky), rather than just sky. I mention it again as it wasn't addressed directly in the authors response, and want to ensure that this comment was not accidentally missed. 

** We prefer to stick with our original approach, partly to be consistent with previous papers in this series and partly because it is unclear what would be the calibration contribution to the uncertainty in the case of an upper limit.

Figure 7: Could also use some additional y-axis labels, as has been made in figs. 2 & 3.

** Done