What I have found,( only by spending more time than is healthy in the company of a stone or two), is that the shadow cast by a central pillar would be a poor fit for a circle. As you say most of the stones would be missed. At midsummer the shadow would follow a kind of parabola shape, like an upturned wine glass, with the midday shadow being almost nothing ( a 6 foot pillar would cast a one foot shadow, in the UK )
At the Equinox's it would almost be a straight line from West to East.
At mid winter it would be the opposite to midsummer.
This is why sundials are not generally pillars. They are normally now a thin slab type structure ( except for some clever fanciful designs) with one edge at least angled to match the tilt of the Earths axis. Rather than the tip of the shadow telling the time, it's where the line cast by the angled slope crosses the hour marks. Even then most modern sundial hour lines won't be laid out truly circular, but more of a parabola shape again.
To get the angle of the slope right it has to match the latitude of the site. Which means it has to point at what is now Polaris, the pole star, or the area in the sky known as the North Celestial Pole.
In the case of Maen Llia this is a slope, pointing true North, of just under 52 deg's. The latitude of Maen Llia is also just under 52 deg's, and the mid summer sunrise 5000 years ago, when viewed from Maen Llia, was also just under 52 deg's. Genius, and just some of the reason why Maen Llia is such a special stone, and probably a better example than Gardom's edge.
This is explained much better on the British sundial society website.