Expecting perfect dots with a 5 minute exposure is asking too much. When you consider the rotation of the galaxies in the sky, you're inherently attempting to translate the movement of a spherical view onto a planar surface. The geometric differences between sphere and plane become far more obvious with extreme wide angles, which is why many users are reporting astrotracer failure using ultrawide 14mm lenses. It's not the astrotracer, it's the geometric translation of the actual world onto the flat sensor... no amount of sensor movement can correct for that for such long exposures.
Here's my proof - I shot this with a K-1 using 1 minute exposures on repeat.
Instagram
Watch the stars exclusively at the left end of the frame. they're not trailing in the conventional usage of the world, but they're moving INTO the frame faster than they're moving when they near the center of the frame. That's the nature of spherical rotation versus simple planar linear transformation. And that's at 24mm. The effect is far more pronounced at 14mm.
The astrotracer gives Pentax abilities that no other camera manufacturer can offer without significant 3rd party equipment. Just like you wouldn't crank the clarity or sharpen sliders up all the way in post, extending the astrotracer exposure time to the max is also not responsible practice. Any other camera manufacturer shooting 1 min exposures at 24mm would have legit trailing all across that timelapse. Pentax astrotracer gives me cleaner dots at 1 min than other manufacturers get at 20 seconds... plus the extra 1.25 stops.