The basic reason is that it’s believed that the likes of the Great Pyramid were arranged to align with particular stars, while Stonehenge and Newgrange were aligned with the Sun.
Why there’s the difference is one of those things that I find is easily understood if you visualise it just right, but it can be a bit difficult to explain with just words. Anyway …
Imagine the Earth going round the Sun over the course of a year. That orbit defines a flat plane with the Sun stationary at the centre. The Earth is also spinning about its own axis once a day. Over the course of the year, this axis is roughly fixed and at the moment happens to point towards the very distant star Polaris. So, although the Earth is going around the Sun, this axis always points towards that star.
So imagine somehow building a pyramid at the North Pole. And include a shaft in it that points directly upwards. You’ve now got a shaft that (roughly) points to the Pole Star throughout the entire year. Job done.
Except that over tens of thousands of years the Earth’s axis isn’t quite fixed. Precession twists it away from the direction of the Pole Star. The angle the axis makes with the plane of the Earth’s orbit remains (roughly) fixed, but precession twists it in a cycle. So our elaborate scheme to incorporate an alignment with Polaris in our new pyramid doesn’t work forever.
That’s the simpler case. Now think what an observer at the Sun sees the Earth’s axis do over the course of a year. Because, over the course of that year, it’s fixed relative to the Pole Star rather than the Sun, they see it twist around. Sometimes they see the North Pole towards them, then they see the axis “side-on”, then they see the North Pole leaning away from them, etc. Over the course of the year, they see the axis twist around entirely.
What does precession look like from the Sun? Although the timescale is much, much longer, it just twists the Earth’s axis through the same cycle. Superimpose the two effects and, apart from details about the timing, the overall effect is that the observer at the Sun just sees the Earth’s axis go through this same cycle each year regardless.
Precession thus doesn’t have any fundamental effect on the cycle of relative geometries the Earth and the Sun undergo over the course of a year. From the point of view of someone on the Earth, precession thus doesn’t change the apparant path of the Sun, but does slowly change the positions of the stars.
Since the most obvious and uncontroversial alignments at Stonehenge and Newgrange are to the Sun, these still work. The Sun does still come up over the Heel Stone at the summer solstice sunrise. And always will.
There are more minor effects that do slightly alter these alignments and people do occasionally try to date the monuments using them. Personally, I usually find that these effects are so small that the astronomical datings for such solar alignments are unconvincing. Using precession to date stellar alignments is usually far more convincing.
Incidentally, the main alignments at Karnak are indeed usually suggested to be solar, but do also still work.