I've never turned on the electronic compass as part of my process with working with it but I don't see how that will help. I have used it to check reasonable alignment when centering Polaris in the frame. I would lean towards it not being a good calibration as it does take some experience to get good ones, or something in the environment affecting it. Also with my recently acquired K-3ii I ended up finding out that the electronic level sensor was horribly miscalibrated (off by almost 20 degrees) on one of the axes so it ended up having no idea how it was actually oriented in space so that would be something to check as well if you haven't. The place I got that camera from has a 6 month warranty on used gear so I brought it back and made it their problem. Several weeks later it arrived back after a trip to Precision Camera and is not properly calibrated.
As far as what can affect it in the environment look for large amounts of iron, so setting up near vehicles, fences, the iron range in northern MN as these will throw it off. Other things that I have found throw it off include electric motors, or anything that has a magnetic field, especially buried power lines or transformers. I've done all of those a number of times once setting up in the dark at the end of a road between 2 transformers out in the brush a little ways with a buried power line right below the tripod, once not thinking and setting up by to the air source heat pump below my deck, a number of times trying to get night shots up by Split Rock lighthouse. Forgetting that one pair of my hunting gloves has a flip open mitten portion that has magnets in it to keep the mitten part open if you want it out of the way for a bit.
Smaller amounts of iron I'm not sure how they affect it but with my setup on
my big wooden tripod with made with regular steel bolts and a Manfrotto 3047 head and hex quick connect plate there doesn't seem to be a problem. I'm not sure how much iron is in the head but it wouldn't surprise me if it were significant given how heavy it is.
Here is a single shot with only very mild quick edits captured with astrotracer and my 400mm wide open for 20 seconds on the setup I mentioned. The lens was still coming down to temp and that was the first quick test shot to check settings, calibration and exposure so was a planned throwaway shot and shows a lot of coma wide open which happens with that giant lens until it reaches thermal equilibrium.
Apart from all of that practice calibrating it with smooth movements along each axis and only one axis at a time. The smoother and more limited to a single axis each movement is the better calibration you end up with. I also have found it useful to go beyond the 180 degree range recommended. Often times the movements I put the camera through are closer to 270 degrees. With the K-3 + O-GPS1 there is an option to do a regular calibration and a precise calibration so I always do the regular one first followed by the precise one. Even with the K-3ii I will turn astrotracer off but have the GPS on so that I can do a regular calibration first when setting up and then switch astrotracer on to do the precise one. I would assume that the K-1 would be similar to the K-3ii in how it functions and is calibrated.
My experience is that the suggested times astrotracer provides are optimistic at best. However at 105mm a 50 or 60 second exposure should be expected to provide good results given that I will regularly do 30 seconds with a 300mm and 20s with a 400mm.