Static palpation is indeed of limited value, probably because of variations in anatomy and poor correlation of supposed findings with pain, dysfunction or pathology. The bone out place theory has not been taught at chiropractic colleges in Europe for 30 years and I don’t use static palpation of bony landmarks for diagnosis at all.
I do however use movement palpation as described in the Gillet test in the references Adam Meakin quotes. Despite poor inter examiner reliability studies I believe motion palpation has a role. Some motion palpation studies have shown better inter-examiner reliability when the design of the study is improved. Although of the cervical spine it shows improved reliability with a different study design. Having palpated an SIJ an estimated 50,000 times over the last 27 years I believe I have built up some skill in assessing these small movements.
As with the other orthopaedic tests Meakin cites for SI examination however, no one test is diagnostic. Motion palpation, taken in combination with the clinical picture, other tests and clinical experience can be helpful in reaching a working diagnosis. Using the correlation of many tests is an approach common to diagnosing many msk conditions.
In addition to motion palpation and orthopaedic tests I find joint line tenderness to be helpful to assist in reaching a conclusion. In addition, prone springing on either side of the SIJ will often reveal decreased give on the hypomobile side as well as tenderness. Some studies have demonstrated altered muscle activity in the presence of SI pain and I often find weakness on testing of rectus femoris, hamstrings or glut max for instance. This may be an effect of arthogenic inhibition but studies have mostly looked at hip and knee pathology in this regard so that is my personal theory to explain some of my clinical findings.