are those coordinates accurate on the image? If they are, it appears your image has the buoy around 105 W, not 149 W. Or am I missing something?
You're right. What you see is what I got taking a screenshot of EOSDIS Worldview display coordinates, it looks like they have a grievous programming error in the mouse-over lat, lon coordinates as well as the snap box corners. The text is so small I could not read the longitude labels on the radial lines to see the conflict.
This is not a confusion of E and W but may be some problem with coordinates not keeping pace with zoom. Worldview does not allow entry/display of coordinates or strings of them (paths) in the manner of Google Earth kml. It offers two distinct lat,lon grids, one simply called 'polar view graticule' and the other EPSG3995 that draw the latitude circles in different places. There might be a slight effect if one used the WGS84 ellipsoid and the other didn't.
Meanwhile nullschool does offer a variety of projections but these seem very sketchily described and not referenced to the international nomenclature standards (WorldView uses EPSG3413). None of them seem to exactly match the plate carree (EPSG: 4326) system of Google Earth which itself has been the source of endless confusion.
Nullschool does not offer any mechanism for snapping to a standard view or returning to a previous one. The grid option provides of unlabelled latitude circles, no longitude lines, no scale, no orientation. I get the sense that the server is still going great but coding stopped a couple years ago.
It would be very convenient to put buoy paths into Google Earth but not so convenient to load WorldView satellite image or nullschool weather data on top, despite an option for doing so. On the other hand, there is no convenient export of vector data out of Google Earth into nullschool or Worldview that I know of. It does seem feasible to get WorldView and nullschool to overlay very well using the 'Stereographic' projection in the latter followed by a scale change and rotation (example a few posts back).
Overall we've run into similar incompatibilities many times before. NSIDC uses the best choice for polar view and a lot of our resources also use that. The Arctic and Antarctic are special situations and for some sites, just a sideline that they have not overly concerned themselves with. Landsat-8 and Sentinel 2A scenes are so tiny in extent compared to the whole Arctic Ocean that the little UTM zones make sense for them.
It is an unfortunate situation that we can't effortlessly marshall all the data in one sensible coord system. Every time something has to be re-projected, the data is degraded. Only a very few operations are harmless.