My feeling is that many of the sites are placed very carefully for viewing rises and settings. Many others however at first sight seem somewhat randomly placed although without closely studying a site it is difficult to tell. For the web site I have generated the views based mainly on the SRTM data but this does not include cairns and other fine details (such as notches) that may have been used.
I suspect some/many sites were placed for reasons other than archaeoastronomy reasons, probably at other areas of interest. The site does provide a really good starting position for studying sites.
At some time I will probably try and run some sort of statistical analysis across the sites to try and classify useful archaeoastronomy sites.
David