R:"The sky annulus had a radius of 12 pixels and was 6 pixels wide, and outliers were removed using a 3-sigma clipping algorithm."
An annulus has two radii, an inner one and an outer one. The text, as written, is ambiguous. The use of 3-sigma as a clipping limit is arbitrary and not statistically supported. If there are billions of data points involved, the laws of statistics would predict multiple 3-sigma outliers, whereas with a handful of data points, 3-sigma outliers would be extremely unlikely. A better approach is to use the Chauvenet criterion, in which the clipping limit is set where the probability of occurrence drops to less than 1/2N, where N is the number of data points. Even that is a somewhat arbitrary limit (1/2N, as opposed to 1/3N, for example), but it is far better than using a fixed 3-sigma limit, independent of the number of data points involved.
A: We put there the adjective "inner" (radius), which removes the ambuiguity.  Then we have added there a few more sentences commenting on the 
bad pixel removal.  As for the reviewer's suggestion for using the Chauvenet criterion, it seems inappropriate for the task of removing bad pixels in the
sky annulus.  We note that its applicability is limited to a data set that has a normal distribution of values, and it should not be used iteratively. 
This is NOT true for the case of sky background, where there is a (probable) normal distribution of counts of "dark" sky background in most pixels 
in the annulus, but with a smaller number of discrepant values skewed to higher values because of background stars or cosmic ray hits.  
The use of a clipped mean is appropriate to remove such outliers.  The only decision is then whether 3 sigma or a somewhat larger multiplier 
is appropriate.  It is not being used to specify which values are "outliers" but just to get an appropriate representative sky background value.
We wouldn't expect any significant difference if the multiplier is 3 or 5, if the process was iterated.

R: Figures: The authors have explained why there are double solid curves in some of the figures, but the figures themselves appear to be unchanged. One of the main reasons for showing data points and model fits in the same figure is to graphically present the remaining differences (i.e. the residuals). When data are folded like they have been in some of these figures and two solid curves are present, the reader is unable to associate which data points correspond to which model fit. Surely a better way can be found to present these data. One option is to show the data unfolded, another would be to use different colors for the folded data.
A: Following the reviewer's suggestion, we have color-coded the data points and the curves fitted to them (outside mutual events) from different nights in Figs. 6-10 now.  While we thought that the quality of the fit to 
the data (outside events) could be assessed reasonably well from the panels b of the figures (where the data points are plotted after subtraction of the primary rotational lightcurve and they effectively correspond 
to fit residuals outside the events), we agree that it is probably useful to present the data also color coded in the plots where they heavily overlap when folded with the orbital period.  We have added one sentence
on it to the caption of Fig. 6.

R: Fourier coefficients: It is this reviewer's opinion that it is unacceptable to delay publication of the coefficients by over a year. The coefficients are available now and should be published now. The authors have admitted that the DART database will include data from the upcoming 2022-2023 apparition, which suggests to me that new fits will be performed to the augmented database, so the coefficients will almost certainly change, or at least will be known to higher precision, so it is entirely possible that the coefficients used to generate the solid curves shown in this paper will never be shown in the DART database. This paper needs to stand on its own. Without the coefficients, it does not.
A: We present the Fourier coefficients in Appendix A now.