This is an interesting study investigating possible orientation strategies used by red knots crossing the Greenland icecap. It provides valuable insights into the tracks of two spring and two autumn migrations of one bird. Comparison of these tracks with the rhumbline and orthodrome shows that the birds follow the orthodrome route more closely, which indicates that the birds may use a time-compensated sunset compass for orientation.
I am missing an evaluation of the possibility of the use of a constant magnetic compass course. In the introduction this possibility is discussed, but there is no effort made to test this hypothesis. Magnetic declination data is easily available online, and I am sure also directly in R, and the routes are easy to model. So I see no reason not to include constant magnetic compass routes in the analysis and add them to the comparison.

Specific comments:
Abstract line 12: I don’t see what is novel here - the use of satellite tracking might be new for birds of the size of knots, but since the method is far from ‘novel’, I would avoid using this word.
Lines 57-59: I would rephrase this sentence and replace the word ‘restrict’. Constant daylight does not ‘restrict’ the use of the sun compass, and the magnetic compass can certainly be used at high latitudes, but the resulting courses might not be very adaptive.
Lines 70-71: Not sure what is meant here…
Line 185: ‘closest to the middlemost time of each migratory bout’: middlemost by definition is the point or time nearest to the exact middle. Since the middle of the migratory bout is an exact time, replace ‘middlemost’ with ‘middle’.
Lines 272-275: The resetting of the clock at a stopover site is not a requirement for the use of a time-compensated sun(set) compass. Birds may also use such a compass based on departure values from the initial departure location or previous stopover sites. See Alerstam and Pettersson 1991, Muheim et al. 2018.
Lines 342-344: This is an anthropocentric view of which compass strategies are ‘simpler’ than others. For an animal orienting at high latitudes, it is probably far more ‘easy’ to follow an orthodrome with a time-compensated sunset compass than to follow a rhumbline for which they would either need access to a star compass (not available) or a sun compass which is constantly time compensated throughout the flight, and not just at departure as is the case with a time-compensated sunset compass.
Fig. 1: I find this figure very difficult to read. The tracks are hardly visible on the coloured background. I suggest to simplify the map by removing the topographic map information, which gives little to no information. Also, the background colour, if any at all, should be chosen to provide an as high contrast to the colours of the tracks as possible. Also consider to use a Mercator projection which is the more classical projection used to illustrate migration routes.