Well adding a study to a site increases the count, so what you suggested doesn't quite work..
if you added a study to site A the count would become 4
adding one to site B, the count there becomes 3.
Makes sense?
Thanks!